A method for constructing an installation accuracy control system for ultra-high-speed maglev supports considering known-point precision Authors: Hongji Xu, Xuefeng Yang, Hang Luo, Liwei Zhang, Chenglong liu Published: 2026 ABSTRACT To meet the extreme measurement accuracy requirements of the Polymorphic Coupled Rail Transit Dynamic Model Test Platform under complex design conditions—namely ultra-high speeds of up to 1,500 km/h and ambient pressures ranging from 0.005 to 1.0 standard atmospheres—this study addresses the lack of a systematic accuracy control theory for installation measurements of ultra-high-speed maglev vacuum-tube supports. As the fundamental positioning elements of the vacuum maglev tube system, support installations directly govern track smoothness and tube airtightness. However, a comprehensive accuracy control framework for this domain has yet to be established, substantially constraining reliable engineering implementation. Here, a “reverse” accuracy allocation strategy is proposed, taking support design parameters as the starting point and integrating the law of error propagation, Least Squares Collocation theory, and computer-based virtual observation simulations. This approach establishes a complete accuracy transfer chain, linking support design accuracy to measurement-control point accuracy, station-setting accuracy, and ultimately to the precision of the offline control network. On this basis, a three-dimensional free-station adjustment model for total station that incorporates the prior accuracy of known points is developed, effectively mitigating the influence of control-point errors on station-setting precision. Through dense grid-based simulation experiments, the spatial distribution characteristics of station-setting accuracy are systematically revealed, providing a theoretical foundation for optimal station placement in field operations. Case studies validate the feasibility and applicability of the proposed methodology, demonstrating that it delivers a rigorously derived accuracy control framework for measurement scheme design under varying tube parameters, instrument precisions, and control network configurations. The accuracy derivation system established in this work offers a complete technical pathway—from theoretical formulation to engineering application—for support installation measurements in China’s ultra-high-speed vacuum maglev transportation systems. It also provides an important reference for promoting the standardized and theory-driven application of precision measurement technologies in major engineering projects. Introduction With the continued advancement of China’s strategy to build a transportation powerhouse, the rail transit sector has achieved sustained and transformative progress, steadily moving toward higher speeds, greater intelligence and improved efficiency. By the end of 2024, the total length of the national railway network had exceeded 160,000 km, of which 45,000 km were high-speed railways (HSR), ranking among the world’s leading systems1. Driven by ongoing technological innovation, the pursuit of ever higher operating speeds in rail transportation has intensified. In this context, maglev rail transit has attracted widespread attention as a key development direction for “future transportation” 2,3. The Polymorphic Coupled Rail Transit Dynamic Model Test Platform, a domestically pioneering and internationally leading innovative research facility, serves as a critical national platform for advancing research on ultra-high-speed vacuum tube transportation in China. It is distinguished by a long experimental track, comprehensive functionality, ultra-high operating speeds, multiple levitation modes, adjustable operating environments and variable model scales. The platform features an elevated test line with a total length of 1.62 km and a tube inner diameter of 3 m, within which the internal air pressure can be continuously regulated over a range of 0.005–1.0 standard atmospheres. With a designed maximum test speed of up to 1,500 km/h, the facility represents an advanced demonstration of “future transportation” technologies2. Ultra-high-speed maglev systems impose exceptionally stringent requirements on track smoothness and vacuum tube sealing, with construction tolerances reaching the millimetre and even sub-millimetre level. As the fundamental components governing track alignment, the installation accuracy of vacuum tube supports directly determines the laying quality and operational stability of the track. Even minor deviations in support positioning can induce significant track geometric deformation under high-speed operating conditions, thereby compromising operational safety and the reliability of experimental data. Achieving high-precision installation measurement of supports is therefore essential. However, a mature technical framework for the installation measurement of maglev supports has yet to be established, and significant methodological gaps and practical challenges remain both domestically and internationally. Owing to its flexibility in station selection and strong adaptability, free-station positioning with a TS has become a core technique in precision engineering surveying, including HSR applications, and vacuum maglev tube support installation is likewise based on this method. Existing studies have mainly focused on the optimisation of free-station geometric configurations and error propagation mechanisms. For example, Horemuz and Andersson examined the effects of the number and planar distribution of common points on station-point accuracy using a trial-and-error approach, and suggested that the optimal strategy is to place most common points near the TS, with only a small number positioned farther away4. Sun et al. simulated dual-backsight free-station experiments under different observation accuracies and confirmed that distance measurement errors exert a greater influence on positioning accuracy than angular measurement errors5. Regarding station-position optimisation, Horemuz and Jansson demonstrated that the optimal horizontal position of a TS should be located at the centroid of the control-point group, and that station position has no significant effect on the planar accuracy of measured object points6. Alizadeh-Khameneh et al. further investigated optimisation in the vertical direction and showed that instrument setup height has only a minor effect on the accuracy of the station-point elevation component7. In addition, the engineering applications of the free-station method have continued to expand. Shevchenko et al. applied it to building stability monitoring8, while Yang and Zou proposed an optimisation and accuracy analysis method for track straightness measurement9. However, most of these studies implicitly adopt the classical assumption that known-point coordinates are error-free, with insufficient attention paid to the propagation mechanism of known-point errors in three-dimensional (3D) space. In the context of HSR track geometric alignment measurement, Liu et al. overcame this limitation by accounting for the influence of prior known-point errors in the adjustment model for track geometry measurements using a track measurement trolley, from the perspective of 3D spatial coordinate transformation10. Although previous studies have laid an important foundation, substantial research gaps remain for maglev tube engineering requiring ultra-high measurement precision. Existing literature has not yet provided a systematic analysis of the coupled mechanism among known-point precision, station-point spatial distribution and station-point accuracy, nor has a 3D free-station optimisation model for TS been established that fully incorporates known-point precision. Given the exceptionally stringent support installation accuracy requirements of vacuum maglev tubes designed for operating speeds of up to 1,500 km/h, even minute error sources cannot be neglected. Addressing these issues is therefore critical for establishing an accuracy control system for maglev support installation measurement. On this basis, this study departs from the classical forward control paradigm in control surveying—namely, progression “from the whole to the part” and “from higher order to lower order”—and instead adopts a reverse-thinking strategy. Starting from the final accuracy requirements imposed by vacuum maglev tubes on support installation, the required accuracy indices for total-station setup and the corresponding measurement-region division scheme are derived in reverse. Specifically, this study first introduces the principle of Least Squares Collocation (LSC) to establish a 3D free-station adjustment model for TS that incorporates the prior precision of known points. Monte Carlo-based virtual simulation is then employed to systematically evaluate the spatial distribution characteristics of station-point accuracy under complex operating conditions. Finally, an integrated accuracy control system is developed, encompassing offline control network configuration, free-station accuracy control, and the determination of installation measurement and control point indices. The proposed method exhibits strong engineering adaptability. It enables dynamic calculation of measurement accuracy indices according to the design parameters of different vacuum maglev tube systems, thereby supporting a more scientific and rational control of accuracy throughout the entire process of maglev support installation. Accuracy indices for support installation measurements According to the design specifications of the Polymorphic Coupled Rail Transit Dynamic Model Test Platform, the installation accuracy design indices for vacuum tube supports are classified into two categories: absolute position parameters and support attitude parameters. The absolute position parameters comprise the longitudinal offset along the bridge axis (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{x}$$\end{document}), the transverse offset across the bridge (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{y}$$\end{document}), and the vertical offset (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{z}$$\end{document}). The support attitude parameters include the rotation about the bridge longitudinal axis (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{x}$$\end{document}), the rotation about the transverse bridge axis (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{y}$$\end{document}), and the rotation about the vertical bridge axis (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{z}$$\end{document}). The corresponding calculation formulations are given below. A 3D coordinate system aligned with the tube axis is first established: a point on the vacuum tube axis is taken as the origin, the tube extension direction defines the x-axis, the direction perpendicular to the tube defines the y-axis, and the z-axis is normal to the x–y plane. All measurements in this study are conducted within this coordinate system. Owing to the specialised manufacturing process of the tube supports, their surfaces do not allow prisms to be mounted directly. Therefore, dedicated measurement fixtures are required in practical measurements. The fixture adopts a drawer-type design, as shown in Fig. 1, and is attached to the support in a prescribed manner. Four high-precision fixed prisms are installed at its corners. In Fig. 1a, P1, P2, P3 and P4 denote the support plate measurement points corresponding to the four prisms under the design state; in Fig. 1b, the light-blue upper plate represents the tube support, and the grey embedded component represents the measurement fixture. With sub-millimetre-level fabrication and assembly accuracy, the fixture ensures both a stable spatial relationship with the support and accurate relative positioning among the prisms. During support installation measurement, the TS directly observes the “spatial virtual plane” formed by these four prisms. Given the sub-millimetre accuracy of the fixture, this virtual plane can be regarded as strictly parallel to the upper surface of the support in space, and the attitude parameters it reflects can therefore directly characterise the actual attitude of the support upper surface. Fig. 1Vacuum tube support and measurement fixture. Vacuum tube support and measurement fixture. As shown in Fig. 1a, the absolute position of a vacuum tube support is referenced to the centre point \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{m}$$\end{document} on its surface. The 3D coordinates of this centre point are determined by averaging the coordinates of the four prisms mounted on the measurement fixture, that is:1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{l}{x}_{{P}_{m}}=\frac{1}{4}({x}_{{P}_{1}}+{x}_{{P}_{2}}+{x}_{{P}_{3}}+{x}_{{P}_{4}})\\\:{y}_{{P}_{m}}=\frac{1}{4}({y}_{{P}_{1}}+{y}_{{P}_{2}}+{y}_{{P}_{3}}+{y}_{{P}_{4}})\\\:{z}_{{P}_{m}}=\frac{1}{4}({z}_{{P}_{1}}+{z}_{{P}_{2}}+{z}_{{P}_{3}}+{z}_{{P}_{4}})-v\end{array}\right.$$\end{document} Where v denotes the relative height difference between the prism centre and the surface of the vacuum tube support, that is, the prism height. The absolute position parameters \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{x}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{y}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{z}$$\end{document} are calculated as follows:2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{c}Dx={x}_{{P}_{m}meas}-{x}_{{P}_{m}des}\\\:Dy={y}_{{P}_{m}meas}-{y}_{{P}_{m}des}\\\:Dz={z}_{{P}_{m}meas}-{z}_{{P}_{m}des}\end{array}\right.$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{x}_{{P}_{m}meas}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{y}_{{P}_{m}meas}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{z}_{{P}_{m}meas}$$\end{document} denotes the measured value, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{x}_{{P}_{m}des}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{y}_{{P}_{m}des}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{z}_{{P}_{m}des}$$\end{document} denotes the design value. As illustrated in Fig. 2, m₁ and m₂ denote the transverse-axis centre points at the large- and small-mileage ends, respectively, while m₃ and m₄ represent the longitudinal-axis centre points on the left and right sides. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{L}_{x}$$\end{document} is the designed longitudinal distance between the measurement and control points of the fixture, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{L}_{y}$$\end{document} is the designed transverse distance between these points. The attitude of the vacuum tube support is controlled using three parameters, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{x}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{y}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{z}$$\end{document}, which are computed as follows:3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{c}Rx={arctan}\frac{\varDelta\:{z}_{{m}_{3}-{m}_{4}}}{\sqrt{\varDelta\:{x}_{{m}_{3}-{m}_{4}}^{2}+\varDelta\:{y}_{{m}_{3}-{m}_{4}}^{2}}}\cdot\:\frac{\pi\:}{18{0}^{\circ\:}}\\\:Ry={arctan}\frac{\varDelta\:{z}_{{m}_{1}-{m}_{2}}}{\sqrt{\varDelta\:{x}_{{m}_{1}-{m}_{2}}^{2}+\varDelta\:{y}_{{m}_{1}-{m}_{2}}^{2}}}\cdot\:\frac{\pi\:}{18{0}^{\circ\:}}\\\:Rz={arcsin}\frac{\varDelta\:{y}_{{m}_{1}-{m}_{2}}}{\sqrt{\varDelta\:{x}_{{m}_{1}-{m}_{2}}^{2}+\varDelta\:{y}_{{m}_{1}-{m}_{2}}^{2}}}\cdot\:\frac{\pi\:}{18{0}^{\circ\:}}\end{array}\right.$$\end{document} Fig. 2Schematic diagram of the fixture axis system. Schematic diagram of the fixture axis system. Back-calculation of measurement and control point accuracy Based on absolute position parameters According to Eq. (2), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{x}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{y}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{z}$$\end{document} represent the limiting errors of the respective coordinate components of the support centre point. In practice, the limiting error in measurements is commonly taken as twice the standard deviation. Accordingly, the designed installation measurement and control accuracy for the support centre point P should be defined as follows:4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{c}{\sigma\:}_{Px}=\frac{1}{2}{D}_{x}\\\:{\sigma\:}_{Py}=\frac{1}{2}{D}_{y}\\\:{\sigma\:}_{Pz}=\frac{1}{2}{D}_{z}\end{array}\right.$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{Px}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{Py}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{Pz}$$\end{document} denote the designed installation measurement and control accuracies of the respective coordinate components of the support centre point. According to Eq. (1), the coordinates of the support centre point are obtained as the arithmetic mean of the coordinates of four independently observed prism points on the measurement fixture. By applying the law of error propagation11,12, the following relationship can be derived:5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{P}=\frac{1}{\sqrt{4}}\sigma\:$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{{P}_{n}}$$\end{document} denotes the standard deviation of each measurement and control point, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{P}$$\end{document} represents the standard deviation of the support centre point. That is, the permissible limit for the measurement accuracy of each measurement and control point should be defined as follows:6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{l}{\sigma\:}_{{P}_{n}x}=2{\sigma\:}_{Px}={D}_{x}\\\:{\sigma\:}_{{P}_{n}y}=2{\sigma\:}_{Py}={D}_{y}\\\:{\sigma\:}_{{P}_{n}z}=2{\sigma\:}_{Pz}={D}_{z}\end{array}\right.$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{{P}_{n}x}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{{P}_{n}y}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{{P}_{n}z}$$\end{document} denote the installation measurement and control accuracies of the respective coordinate components of the measurement and control points on the fixture. Based on the above expressions, the design values of the absolute position parameters \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{x}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{y}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{z}$$\end{document} for the vacuum tube supports can be converted into the corresponding measurement accuracy indices of the measurement and control points on the fixture, thereby providing the basis for subsequent analyses. Based on support attitude parameters As shown in Fig. 3, panels (a), (b) and (c) respectively illustrate the attitude of the virtual plane formed by the centres of the four prisms on the measurement fixture when only the effects of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{x}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{y}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{z}$$\end{document} are considered. Based on the fixture characteristics described above, the spatial attitude of this virtual plane can be regarded as directly equivalent to the actual attitude of the upper plate surface of the tube support. Fig. 3Schematic illustration of the support upper plate attitude under error effects. Schematic illustration of the support upper plate attitude under error effects. When only the effect of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{x}$$\end{document} is considered, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:{x}_{{m}_{3}-{m}_{4}}=0$$\end{document}. Substituting the actual design dimensions of the support yields:7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\sqrt{\varDelta\:{x}_{{m}_{3}-{m}_{4}}^{2}+\varDelta\:{y}_{{m}_{3}-{m}_{4}}^{2}}=\varDelta\:{y}_{{m}_{3}-{m}_{4}}=Ly\cdot\:{cos}(Rx\cdot\:\frac{180^\circ\:}{\pi\:})$$\end{document} When only the effect of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{y}$$\end{document} is considered, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:{y}_{{m}_{1}-{m}_{2}}=0$$\end{document}. Substituting the actual design dimensions of the support yields:8\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\sqrt{\varDelta\:{x}_{{m}_{1}-{m}_{2}}^{2}+\varDelta\:{y}_{{m}_{1}-{m}_{2}}^{2}}=\varDelta\:{x}_{{m}_{1}-{m}_{2}}=Lx\cdot\:{cos}(Ry\cdot\:\frac{180^\circ\:}{\pi\:})$$\end{document} When only the effect of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{z}$$\end{document} is considered and the actual design dimensions of the support are substituted, the following expression is obtained:9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\sqrt{\varDelta\:{x}_{{m}_{1}-{m}_{2}}^{2}+\varDelta\:{y}_{{m}_{1}-{m}_{2}}^{2}}=Lx$$\end{document} By substituting Eqs. (7), (8) and (9) into Eq. (3) and performing the necessary simplifications, the following result is obtained:10\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{c}\varDelta\:{z}_{{m}_{3}-{m}_{4}}=Ly\cdot\:{sin}(Rx\cdot\:\frac{180^\circ\:}{\pi\:})\\\:\varDelta\:{z}_{{m}_{1}-{m}_{2}}=Lx\cdot\:{sin}(Ry\cdot\:\frac{180^\circ\:}{\pi\:})\\\:\varDelta\:{y}_{{m}_{1}-{m}_{2}}=Lx\cdot\:{sin}(Rz\cdot\:\frac{18{0}^{\circ\:}}{\pi\:})\end{array}\right.$$\end{document} Using the above expression, the designed attitude parameters of the plate surface, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{x}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{y}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{z}$$\end{document}, can be converted into the corresponding incremental limiting values. Taking twice the standard deviation as the limiting criterion, the following relationship is obtained:11\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{c}{\sigma\:}_{\varDelta\:{z}_{{m}_{3}-{m}_{4}}}=\frac{Ly}{2}\cdot\:{sin}(Rx\cdot\:\frac{180^\circ\:}{\pi\:})\\\:{\sigma\:}_{\varDelta\:{z}_{{m}_{1}-{m}_{2}}}=\frac{Lx}{2}\cdot\:{sin}(Ry\cdot\:\frac{180^\circ\:}{\pi\:})\\\:{\sigma\:}_{\varDelta\:{y}_{{m}_{1}-{m}_{2}}}=\frac{Lx}{2}\cdot\:{sin}(Rz\cdot\:\frac{18{0}^{\circ\:}}{\pi\:})\end{array}\right.$$\end{document} Using the coordinate increment calculation formula, the following can be obtained:12\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}\Delta z_{m_3-m_4} &= z_{m_3}-z_{m_4} \\&= \frac{1}{2}\left(z_{P_1}+z_{P_2}\right)-\frac{1}{2}\left(z_{P_3}+z_{P_4}\right) \\&= \frac{1}{2}\left(z_{P_1}-z_{P_4}\right)+\frac{1}{2}\left(z_{P_2}-z_{P_3}\right) \\&= \frac{1}{2}\left(\Delta z_{P_1-P_4}+\Delta z_{P_2-P_3}\right)\end{aligned}$$\end{document}13\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}\Delta z_{m_1-m_2} &= z_{m_1} - z_{m_2} \\&= \frac{1}{2}(z_{P_1} + z_{P_4}) - \frac{1}{2}(z_{P_2} + z_{P_3}) \\&= \frac{1}{2}(z_{P_1} - z_{P_2}) + \frac{1}{2}(z_{P_4} - z_{P_3}) \\&= \frac{1}{2}(\Delta z_{P_1-P_2} + \Delta z_{P_4-P_3})\end{aligned}$$\end{document}14\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}\Delta y_{m_1-m_2} &= y_{m_1} - y_{m_2} \\&= \frac{1}{2}(y_{P_1} + y_{P_4}) - \frac{1}{2}(y_{P_2} + y_{P_3}) \\&= \frac{1}{2}(y_{P_1} - y_{P_2}) + \frac{1}{2}(y_{P_4} - y_{P_3}) \\&= \frac{1}{2}(\Delta y_{P_1-P_2} + \Delta y_{P_4-P_3})\end{aligned}$$\end{document} According to the law of error propagation11,12, the following relationship holds:15\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{c}{\sigma\:}_{\varDelta\:{z}_{{m}_{3}-{m}_{4}}}=\frac{{\sigma\:}_{\varDelta\:{z}_{{P}_{1}-{P}_{4}}}}{\sqrt{2}}=\frac{{\sigma\:}_{\varDelta\:{z}_{{P}_{2}-{P}_{3}}}}{\sqrt{2}}\\\:{\sigma\:}_{\varDelta\:{z}_{{m}_{1}-{m}_{2}}}=\frac{{\sigma\:}_{\varDelta\:{z}_{{P}_{1}-{P}_{2}}}}{\sqrt{2}}=\frac{{\sigma\:}_{\varDelta\:{z}_{{P}_{4}-{P}_{3}}}}{\sqrt{2}}\\\:{\sigma\:}_{\varDelta\:{y}_{{m}_{1}-{m}_{2}}}=\frac{{\sigma\:}_{{\varDelta\:y}_{P1-P2}}}{\sqrt{2}}=\frac{{\sigma\:}_{{\varDelta\:y}_{P4-P3}}}{\sqrt{2}}\end{array}\right.$$\end{document} By substituting the above expressions into Eq. (11) and simplifying, the following result is obtained:16\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{c}{\sigma\:}_{\varDelta\:{z}_{{P}_{1}-{P}_{4}}}={\sigma\:}_{\varDelta\:{z}_{{P}_{2}-{P}_{3}}}=\frac{Ly}{\sqrt{2}}\cdot\:{sin}(Rx\cdot\:\frac{180^\circ\:}{\pi\:})\\\:{\sigma\:}_{\varDelta\:{z}_{{P}_{1}-{P}_{2}}}={\sigma\:}_{\varDelta\:{z}_{{P}_{4}-{P}_{3}}}=\frac{Lx}{\sqrt{2}}\cdot\:{sin}(Ry\cdot\:\frac{180^\circ\:}{\pi\:})\\\:{\sigma\:}_{\varDelta\:{y}_{{P}_{1}-{P}_{2}}}={\sigma\:}_{\varDelta\:{y}_{{P}_{4}-{P}_{3}}}=\frac{Lx}{\sqrt{2}}\cdot\:{sin}(Rz\cdot\:\frac{180^\circ\:}{\pi\:})\end{array}\right.$$\end{document} Using the above expression, the limiting accuracy requirements for the designed plate attitude can be transformed into corresponding limits on the accuracy of the coordinate increments of the prism measurement and control points on the measurement fixture. Back-calculation of free-station point accuracy Back-calculation based on absolute position parameters Based on the method described above, the design values of the absolute position parameters of the vacuum tube supports (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{x}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{y}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{z}$$\end{document}) can be transformed into the corresponding independent measurement accuracy indices (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{x}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{y}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{z}$$\end{document}) for each measurement and control point on the measurement fixture. During the installation measurement of ultra-high-speed vacuum maglev tube supports, a free-station TS configuration is adopted. Consequently, the installation measurement of an individual measurement and control point is essentially realised through 3D free-station positioning combined with polar-coordinate stake-out using a TS, as illustrated in Fig. 4. The stake-out point corresponds directly to the measurement and control point on the installation fixture. Fig. 4Schematic illustration of stake-out using the polar-coordinate method. Schematic illustration of stake-out using the polar-coordinate method. According to the coordinate increment calculation formula, the coordinates of the stake-out point can be expressed as13:17\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{l}{x}_{p}={x}_{s}+D\cdot\:{cos}{\alpha\:}_{P}\\\:{y}_{p}={y}_{s}+D\cdot\:{sin}{\alpha\:}_{P}\\\:{z}_{p}={z}_{s}+S\cdot\:{cos}A\end{array}\right.$$\end{document} Where, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$D=S\cdot{sin}A$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\alpha}_{P}=\omega\:+\alpha\:\pm\:36{0}^{\circ}$$\end{document}, then:18\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{l}{x}_{p}={x}_{s}+S\cdot\:{sin}A\cdot\:{cos}(\omega\:+\alpha\:)\\\:{y}_{p}={y}_{s}+S\cdot\:{sin}A\cdot\:{sin}(\omega\:+\alpha\:)\\\:{z}_{p}={z}_{s}+S\cdot\:{cos}A\end{array}\right.$$\end{document} By successively taking the total differentials of the above expression, the following result is obtained:19\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{rcl}d{x}_{p} &=& \left(\frac{\partial f}{\partial {x}_{s}}\right)d{x}_{s}+\left(\frac{\partial f}{\partial \omega}\right)\cdot\frac{1}{\rho}\cdot d\omega+\left(\frac{\partial f}{\partial S}\right)dS+\left(\frac{\partial f}{\partial A}\right)\cdot\frac{1}{\rho}\cdot dA+\left(\frac{\partial f}{\partial \alpha}\right)\cdot\frac{1}{\rho}\cdot d\alpha \\&=& d{x}_{s}-\varDelta {y}_{SP}\frac{1}{\rho}d\omega+\sin A\cdot \cos{\alpha}_{P}dS+\varDelta {z}_{SP}\cos{\alpha}_{P}\frac{1}{\rho}dA-\varDelta {y}_{SP}\frac{1}{\rho}d\alpha\end{array}$$\end{document}20\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{rcl}d{y}_{p} &=& \left(\frac{\partial f}{\partial {y}_{s}}\right)d{y}_{s}+\left(\frac{\partial f}{\partial \omega}\right)\cdot\frac{1}{\rho}\cdot d\omega+\left(\frac{\partial f}{\partial S}\right)dS+\left(\frac{\partial f}{\partial A}\right)\cdot\frac{1}{\rho}\cdot dA+\left(\frac{\partial f}{\partial \alpha}\right)\cdot\frac{1}{\rho}\cdot d\alpha \\&=& d{y}_{s}+\varDelta{x}_{SP}\frac{1}{\rho}d\omega+\sin A\sin{\alpha}_{P}dS+\varDelta{z}_{SP}\sin{\alpha}_{P}\frac{1}{\rho}dA+\varDelta{x}_{SP}\frac{1}{\rho}d\alpha\end{array}$$\end{document}21\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{rcl}d{z}_{p} &=& \left(\frac{\partial f}{\partial {z}_{s}}\right)d{z}_{s}+\left(\frac{\partial f}{\partial S}\right)dS+\left(\frac{\partial f}{\partial A}\right)\cdot\frac{1}{\rho}\cdot dA \\&=& d{z}_{s}+\cos A\cdot dS-S\sin A\cdot\frac{1}{\rho}\cdot dA\end{array}$$\end{document} Rearranging the above expression and letting:22\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:dx={\left[\begin{array}{cccc}d{x}_{s}&\:d{y}_{s}&\:d{z}_{s}&\:d\omega\:\end{array}\right]}^{T}$$\end{document}23\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:ds=\left[\begin{array}{ccc}dS&\:dA&\:d\alpha\:\end{array}\right]$$\end{document}24\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{F}_{\alpha\:}=\left[\begin{array}{cccc}1&\:0&\:0&\:-\frac{\varDelta\:{y}_{SP}}{\rho\:}\\\:0&\:1&\:0&\:\frac{\varDelta\:{x}_{SP}}{\rho\:}\\\:0&\:0&\:1&\:0\end{array}\right]$$\end{document}25\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{F}_{s}=\left[\begin{array}{ccc}{sin}A\cdot\:{cos}{\alpha\:}_{P}&\:\varDelta\:{z}_{SP}{cos}{\alpha\:}_{P}\frac{1}{\rho\:}&\:-\varDelta\:{y}_{SP}\frac{1}{\rho\:}\\\:{sin}A{sin}{\alpha\:}_{P}&\:\varDelta\:{z}_{SP}{sin}{\alpha\:}_{P}\frac{1}{\rho\:}&\:\varDelta\:{x}_{SP}\frac{1}{\rho\:}\\\:{cos}A&\:-S{sin}A\cdot\:\frac{1}{\rho\:}&\:0\end{array}\right]$$\end{document} Then:26\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{3\times\:1}{\left[\begin{array}{c}d{x}_{p}\\\:d{y}_{p}\\\:d{z}_{p}\end{array}\right]}=\underset{3\times\:3}{{F}_{s}}\cdot\:\underset{3\times\:1}{\left[\begin{array}{c}d{S}_{1}\\\:dA\\\:d\alpha\:\end{array}\right]}+\underset{3\times\:4}{{F}_{\alpha\:}}\cdot\:\underset{4\times\:1}{\left[\begin{array}{c}d{x}_{S}\\\:d{y}_{S}\\\:d{z}_{S}\\\:d\omega\:\end{array}\right]}$$\end{document} According to the law of error propagation12,13:27\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{P}={F}_{S}\cdot\:D\cdot\:{F}_{S}^{T}+{F}_{\alpha\:}\cdot\:{D}_{S}\cdot\:{F}_{\alpha\:}^{T}$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{P}$$\end{document} denotes the variance–covariance matrix of the stake-out point P, D represents the variance matrix of the observations, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{S}$$\end{document} denotes the variance–covariance matrix of the station point S. That is:28\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{3\times\:3}{{D}_{P}=\left[\begin{array}{ccc}{{\sigma\:}_{{x}_{P}x}^{2}}_{P}&\:{{\sigma\:}_{{x}_{P}y}^{2}}_{P}&\:{{\sigma\:}_{{x}_{P}z}^{2}}_{P}\\\:{{\sigma\:}_{{y}_{P}x}^{2}}_{P}&\:{{\sigma\:}_{{y}_{P}y}^{2}}_{P}&\:{{\sigma\:}_{{y}_{P}z}^{2}}_{P}\\\:{{\sigma\:}_{{z}_{P}x}^{2}}_{P}&\:{{\sigma\:}_{{z}_{P}y}^{2}}_{P}&\:{{\sigma\:}_{{z}_{P}z}^{2}}_{P}\end{array}\right]}=\underset{3\times\:3}{{F}_{S}}\cdot\:\underset{3\times\:3}{\left[\begin{array}{ccc}{\sigma\:}_{S}^{2}&\:&\:\\\:&\:{\sigma\:}_{A}^{2}&\:\\\:&\:&\:{\sigma\:}_{\alpha\:}^{2}\end{array}\right]\cdot\:}{\underset{3\times\:3}{{F}_{S}}}^{T}+\underset{3\times\:4}{{F}_{\alpha\:}}\cdot\:\underset{4\times\:4}{{D}_{S}}\cdot\:\underset{4\times\:3}{{F}_{\alpha\:}^{T}}$$\end{document} After calculation and simplification, the diagonal elements of the matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{P}$$\end{document} are obtained as follows:29\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{l}{\sigma\:}_{{x}_{P}}^{2}=({sin}A\cdot\:{cos}{\alpha\:}_{p}{)}^{2}{\sigma\:}_{S}^{2}+(\varDelta\:{z}_{SP}\cdot\:{cos}{\alpha\:}_{P}\cdot\:\frac{1}{\rho\:}{)}^{2}{\sigma\:}_{A}^{2}+(\varDelta\:{y}_{SP}\cdot\:\frac{1}{\rho\:}{)}^{2}{\sigma\:}_{\alpha\:}^{2}+{\sigma\:}_{{x}_{S}}^{2}-2\varDelta\:{y}_{SP}\frac{1}{\rho\:}\cdot\:{\sigma\:}_{\omega\:{x}_{S}}^{2}+(\varDelta\:{y}_{SP}\cdot\:\frac{1}{\rho\:}{)}^{2}{\sigma\:}_{\omega\:}^{2}\\\:{\sigma\:}_{{y}_{P}}^{2}=({sin}A\cdot\:{sin}{\alpha\:}_{p}{)}^{2}{\sigma\:}_{S}^{2}+(\varDelta\:{z}_{SP}\cdot\:{sin}{\alpha\:}_{P}\cdot\:\frac{1}{\rho\:}{)}^{2}{\sigma\:}_{A}^{2}+(\varDelta\:{x}_{SP}\cdot\:\frac{1}{\rho\:}{)}^{2}{\sigma\:}_{\alpha\:}^{2}+{\sigma\:}_{{y}_{S}}^{2}+2\varDelta\:{x}_{SP}\frac{1}{\rho\:}\cdot\:{\sigma\:}_{\omega\:{y}_{S}}^{2}+(\varDelta\:{x}_{SP}\cdot\:\frac{1}{\rho\:}{)}^{2}{\sigma\:}_{\omega\:}^{2}\\\:{\sigma\:}_{{z}_{P}}^{2}=({cos}A{)}^{2}{\sigma\:}_{S}^{2}+(S\cdot\:{sin}A\cdot\:\frac{1}{\rho\:}{)}^{2}{\sigma\:}_{A}^{2}+{\sigma\:}_{{z}_{S}}^{2}\end{array}\right.$$\end{document} As indicated by the above expression, the accuracy of the measurement and control points is jointly determined by the observation accuracy and the accuracy of the station point. Back-calculation based on support attitude parameters As demonstrated by the foregoing analysis, the designed support attitude parameters can be converted into accuracy requirements for the coordinate increments \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{\varDelta\:{z}_{{P}_{1}-{P}_{4}}}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{\varDelta\:{z}_{{P}_{1}-{P}_{2}}}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{\varDelta\:{y}_{{P}_{1}-{P}_{2}}}$$\end{document} of the measurement and control points. Consequently, precise control of the coordinate increment accuracy between measurement and control points enables effective control of the support installation attitude. Figure 5 presents a schematic illustration of staking out two measurement and control points using the polar-coordinate method with a free-station TS. Fig. 5Schematic illustration of stake-out using the polar-coordinate method (two measurement and control points). Schematic illustration of stake-out using the polar-coordinate method (two measurement and control points). The coordinates of the stake-out points can be computed using Eq. (18) presented in the preceding section. Accordingly, the coordinate increments between the two stake-out points are given by:30\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{l}\varDelta\:{x}_{{p}_{1-2}}={x}_{{p}_{2}}-{x}_{{p}_{1}}={S}_{2}\cdot\:{sin}{A}_{2}\cdot\:{cos}(\omega\:+{\alpha\:}_{2})-{S}_{1}\cdot\:{sin}{A}_{1}\cdot\:{cos}(\omega\:+{\alpha\:}_{1})\\\:\varDelta\:{y}_{{p}_{1-2}}={y}_{{p}_{2}}-{y}_{{p}_{1}}={S}_{2}\cdot\:{sin}{A}_{2}\cdot\:{sin}(\omega\:+{\alpha\:}_{2})-{S}_{1}\cdot\:{sin}{A}_{1}\cdot\:{sin}(\omega\:+{\alpha\:}_{1})\\\:\varDelta\:{z}_{{p}_{1-2}}={z}_{{p}_{2}}-{z}_{{p}_{1}}={S}_{2}\cdot\:{cos}{A}_{2}-{S}_{1}\cdot\:{cos}{A}_{1}\end{array}\right.$$\end{document} By successively taking the total differentials of the above expression:31\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:{x}_{{p}_{1-2}}=f({S}_{1},{A}_{1},{\alpha\:}_{1},{S}_{2},{A}_{2},{\alpha\:}_{2},\omega\:)$$\end{document}32\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{rcl}d\varDelta{x}_{p_{1-2}}&=&\left(\frac{\partial f}{\partial S_{1}}\right)dS_{1}+\left(\frac{\partial f}{\partial A_{1}}\right)\frac{dA_{1}}{\rho}+\left(\frac{\partial f}{\partial \alpha_{1}}\right)\frac{d\alpha_{1}}{\rho}+\left(\frac{\partial f}{\partial S_{2}}\right)dS_{2}+\left(\frac{\partial f}{\partial A_{2}}\right)\frac{dA_{2}}{\rho}+\left(\frac{\partial f}{\partial \alpha_{2}}\right)\frac{d\alpha_{2}}{\rho}+\left(\frac{\partial f}{\partial \omega}\right)\frac{d\omega}{\rho}\\&=&-\sin A_{1}\cdot\cos(\omega+\alpha_{1})dS_{1}-S_{1}\cos(\omega+\alpha_{1})\cos A_{1}\frac{dA_{1}}{\rho}+S_{1}\sin A_{1}\sin(\omega+\alpha_{1})\frac{d\alpha_{1}}{\rho}\\&&+\sin A_{2}\cdot\cos(\omega+\alpha_{2})dS_{2}+S_{2}\cos(\omega+\alpha_{2})\cos A_{2}\frac{dA_{2}}{\rho}-S_{2}\sin A_{2}\sin(\omega+\alpha_{2})\frac{d\alpha_{2}}{\rho}\\&&+\left[S_{1}\sin A_{1}\sin(\omega+\alpha_{1})-S_{2}\sin A_{2}\sin(\omega+\alpha_{2})\right]\frac{d\omega}{\rho}\end{array}$$\end{document}33\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:{y}_{{p}_{1-2}}=f({S}_{1},{A}_{1},{\alpha\:}_{1},{S}_{2},{A}_{2},{\alpha\:}_{2},\omega\:)$$\end{document}34\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{rcl}d\varDelta{y}_{p_{1-2}}&=&\left(\frac{\partial f}{\partial S_{1}}\right)dS_{1}+\left(\frac{\partial f}{\partial A_{1}}\right)\frac{dA_{1}}{\rho}+\left(\frac{\partial f}{\partial \alpha_{1}}\right)\frac{d\alpha_{1}}{\rho}+\left(\frac{\partial f}{\partial S_{2}}\right)dS_{2}+\left(\frac{\partial f}{\partial A_{2}}\right)\frac{dA_{2}}{\rho}+\left(\frac{\partial f}{\partial \alpha_{2}}\right)\frac{d\alpha_{2}}{\rho}+\left(\frac{\partial f}{\partial \omega}\right)\frac{d\omega}{\rho}\\&=&-\sin A_{1}\cdot\sin(\omega+\alpha_{1})dS_{1}-S_{1}\sin(\omega+\alpha_{1})\cdot\cos A_{1}\frac{dA_{1}}{\rho}-S_{1}\sin A_{1}\cdot\cos(\omega+\alpha_{1})\frac{d\alpha_{1}}{\rho}\\&&+\sin A_{2}\cdot\sin(\omega+\alpha_{2})dS_{2}+S_{2}\sin(\omega+\alpha_{2})\cdot\cos A_{2}\frac{dA_{2}}{\rho}+S_{2}\sin A_{2}\cos(\omega+\alpha_{2})\frac{d\alpha_{2}}{\rho}\\&&+\left[S_{2}\sin A_{2}\cdot\cos(\omega+\alpha_{2})-S_{1}\sin A_{1}\cdot\cos(\omega+\alpha_{1})\right]\frac{d\omega}{\rho}\end{array}$$\end{document}35\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:{z}_{{p}_{1-2}}=f({S}_{1},{A}_{1},{S}_{2},{A}_{2})$$\end{document}36\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{rcl}d\varDelta{z}_{p_{1-2}}&=&\left(\frac{\partial f}{\partial S_{1}}\right)dS_{1}+\left(\frac{\partial f}{\partial A_{1}}\right)\frac{dA_{1}}{\rho}+\left(\frac{\partial f}{\partial S_{2}}\right)dS_{2}+\left(\frac{\partial f}{\partial A_{2}}\right)\frac{dA_{2}}{\rho}\\&=&-\cos A_{1}dS_{1}+S_{1}\sin A_{1}\cdot\frac{dA_{1}}{\rho}+\cos A_{2}\cdot dS_{2}-S_{2}\sin A_{2}\cdot\frac{dA_{2}}{\rho}\end{array}$$\end{document} By rearranging Eqs. (32), (34) and (36), and letting:37\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:dP={\left[\begin{array}{ccccccc}d{S}_{1}&\:d{A}_{1}&\:d{\alpha\:}_{1}&\:d{S}_{2}&\:d{A}_{2}&\:d{\alpha\:}_{2}&\:d\omega\:\end{array}\right]}^{T}$$\end{document}38\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:d\varDelta\:={\left[\begin{array}{ccc}d\varDelta\:{x}_{{P}_{1-2}}&\:d\varDelta\:{y}_{{P}_{1-2}}&\:d\varDelta\:{z}_{{P}_{1-2}}\end{array}\right]}^{T}$$\end{document} It can be expressed as:39\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:d\varDelta\:=K\cdot\:dP$$\end{document} Where K denotes the coefficient matrix. Let the variance matrix of the observations be defined as:40\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:D=diag\left(\begin{array}{ccccccc}{\sigma\:}_{{S}_{1}}^{2}&\:{\sigma\:}_{{A}_{1}}^{2}&\:{\sigma\:}_{{\alpha\:}_{1}}^{2}&\:{\sigma\:}_{{S}_{2}}^{2}&\:{\sigma\:}_{{A}_{2}}^{2}&\:{\sigma\:}_{{\alpha\:}_{2}}^{2}&\:{\sigma\:}_{\omega\:}^{2}\end{array}\right)$$\end{document} Considering the independence of the individual observations, and according to the law of error propagation \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{\varDelta\:}=K\cdot\:D\cdot\:{K}^{T}$$\end{document}, the variance–covariance matrix of the coordinate increments is computed. Taking its diagonal elements yields:41\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left\{\begin{array}{rcl}{\sigma}_{\varDelta x}^{2}&=&\left[\sin A_{1}\cos\left(\omega+\alpha_{1}\right)\right]^{2}{\sigma}_{S_{1}}^{2}+\left[S_{1}\cdot\cos\left(\omega+\alpha_{1}\right)\cos A_{1}\cdot\frac{1}{\rho}\right]^{2}{\sigma}_{A_{1}}^{2}+\left[S_{1}\cdot\sin A_{1}\sin\left(\omega+\alpha_{1}\right)\cdot\frac{1}{\rho}\right]^{2}{\sigma}_{\alpha_{1}}^{2}\\&&+\left[\sin A_{2}\cos\left(\omega+\alpha_{2}\right)\right]^{2}{\sigma}_{S_{2}}^{2}+\left[S_{2}\cdot\cos\left(\omega+\alpha_{2}\right)\cos A_{2}\cdot\frac{1}{\rho}\right]^{2}{\sigma}_{A_{2}}^{2}+\left[S_{2}\cdot\sin A_{2}\sin\left(\omega+\alpha_{2}\right)\cdot\frac{1}{\rho}\right]^{2}{\sigma}_{\alpha_{2}}^{2}\\&&+\left\{\left[S_{1}\cdot\sin A_{1}\sin\left(\omega+\alpha_{1}\right)-S_{2}\cdot\sin A_{2}\sin\left(\omega+\alpha_{2}\right)\right]\frac{1}{\rho}\right\}^{2}\cdot{\sigma}_{\omega}^{2}\\{\sigma}_{\varDelta y}^{2}&=&\left[\sin A_{1}\sin\left(\omega+\alpha_{1}\right)\right]^{2}{\sigma}_{S_{1}}^{2}+\left[S_{1}\cdot\sin\left(\omega+\alpha_{1}\right)\cos A_{1}\cdot\frac{1}{\rho}\right]^{2}{\sigma}_{A_{1}}^{2}+\left[S_{1}\cdot\sin A_{1}\cos\left(\omega+\alpha_{1}\right)\cdot\frac{1}{\rho}\right]^{2}{\sigma}_{\alpha_{1}}^{2}\\&&+\left[\sin A_{2}\sin\left(\omega+\alpha_{2}\right)\right]^{2}{\sigma}_{S_{2}}^{2}+\left[S_{2}\cdot\sin\left(\omega+\alpha_{2}\right)\cos A_{2}\cdot\frac{1}{\rho}\right]^{2}{\sigma}_{A_{2}}^{2}+\left[S_{2}\cdot\sin A_{2}\cos\left(\omega+\alpha_{2}\right)\cdot\frac{1}{\rho}\right]^{2}{\sigma}_{\alpha_{2}}^{2}\\&&+\left\{\left[S_{2}\cdot\sin A_{2}\cos\left(\omega+\alpha_{2}\right)-S_{1}\cdot\sin A_{1}\cos\left(\omega+\alpha_{1}\right)\right]\frac{1}{\rho}\right\}^{2}{\sigma}_{\omega}^{2}\\{\sigma}_{\varDelta z}^{2}&=&\left(\cos A_{1}\right)^{2}{\sigma}_{S_{1}}^{2}+\left(S_{1}\cdot\sin A_{1}\cdot\frac{1}{\rho}\right)^{2}{\sigma}_{A_{1}}^{2}+\left(\cos A_{2}\right)^{2}{\sigma}_{S_{2}}^{2}+\left(S_{2}\cdot\sin A_{2}\cdot\frac{1}{\rho}\right)^{2}{\sigma}_{A_{2}}^{2}\end{array}\right.$$\end{document} According to the accuracy evaluation formula in Eq. (41), the accuracy of the x- and y-coordinate increments between the two stake-out points depends solely on the measurement accuracy and the accuracy of the station orientation angle, whereas the accuracy of the z-coordinate increment depends only on the measurement accuracy; none of these are affected by the coordinate accuracy of the station point. From a macroscopic perspective, as illustrated in Fig. 6, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\omega\:}^{{\prime\:}}$$\end{document} denotes the orientation direction containing an error, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{P}_{1}}^{{\prime\:}}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{P}_{2}}^{{\prime\:}}$$\end{document} represent the two stake-out points influenced by \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\omega\:}^{{\prime\:}}$$\end{document}. Under ideal conditions in which observation errors are neglected, both the lengths and horizontal directions of the two observation sides \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:S{P}_{1}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:S{P}_{2}$$\end{document} are fixed quantities, and consequently the horizontal distance \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{{P}_{1}{P}_{2}}$$\end{document} between points \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{1}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{2}$$\end{document} remains constant. Fig. 6Schematic illustration of the effect of orientation angle errors in polar-coordinate stake-out. Schematic illustration of the effect of orientation angle errors in polar-coordinate stake-out. When an error exists in the orientation angle of the station point S, the relative geometric relationship between the two observation sides remains unchanged; instead, the configuration rotates as a whole about S, with the direction and magnitude of rotation corresponding to the orientation angle error. In this process, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{{P}_{1}{P}_{2}}$$\end{document} remains constant and, together with \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:x$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:y$$\end{document}, forms a right triangle, leading to variations in \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:x$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:y$$\end{document}. Consequently, the accuracy of the coordinate increments between the two stake-out points depends not only on the observation accuracy but also on the accuracy of the station orientation angle. Although the orientation angle parameter enters the coordinate computation as a constant, it is in fact a random variable with known precision. Therefore, the orientation angle accuracy \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{\omega\:}$$\end{document} of the station point must also be treated as an accuracy evaluation index. In summary, by substituting the measurement and control point accuracy requirements derived from the support design parameters into Eqs. (29) and (41), the theoretical weakest coordinate accuracy and orientation angle accuracy required for the station point can be determined for instruments of different accuracy classes, while simultaneously considering the constraints imposed by both the absolute position parameters and attitude parameters of the support. This achieves the objective of the “reverse” accuracy allocation method proposed in this study: starting from the actual support installation accuracy requirements of vacuum maglev tubes and inversely deriving the required station-point accuracy with precision. The resulting framework enables the actual accuracy demand to be rigorously controlled and provides a solid theoretical basis for station-point accuracy design. It also ensures that the specified accuracy indices are scientifically and rationally defined, thereby avoiding both the risk of non-compliant construction caused by insufficient accuracy requirements and the reduction in construction efficiency or accuracy redundancy caused by overly stringent requirements. A 3D free-station adjustment model for total station based on the principle of Least Squares Collocation In HSR engineering, control points are commonly treated as error-free, non-random parameters in adjustment computations, an assumption that neglects the inevitable errors present in practice. For vacuum maglev tube systems with design operating speeds of up to 1,500 km/h, however, the installation accuracy requirements far exceed those of conventional engineering projects, and the influence of control point errors cannot be ignored. Moreover, the offline control points used in free-station positioning are themselves derived from adjustments of higher-order control networks and therefore possess prior accuracy information. They should thus be treated as random parameters with known prior precisions in the adjustment process. To properly account for the influence of control point accuracy, this study proposes a 3D free-station method for TS based on the principle of LSC thereby enhancing the reliability and precision of measurement and control in vacuum maglev tube support installation. The principle of LSC is an adjustment approach that introduces random parameters (signals) with known prior expectations and variances into the functional model, while simultaneously accounting for both observation error variances and signal variances in accuracy assessment14. Its functional model is given by14:42\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:L=AX+BY+\varDelta\:$$\end{document} Where X denotes the non-random parameters (deterministic parameters), Y represents the random parameters (signals), L denotes the observations, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:$$\end{document} corresponds to the noise, that is, the observation errors. The corresponding error equation is given by:43\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:V=A\cdot\:dX+B\cdot\:dY-L$$\end{document} Where V denotes the vector of observation corrections, dX is the vector of corrections to the parameters, and dY is the vector of corrections to the signals. Computation of approximate station-point coordinates In the 3D free-station method for TS based on the LSC principle, the original observations consist of the slope distances, horizontal directions and zenith distances between the station point and each control point. The station-point coordinates and orientation angle are treated as non-random parameters (parameters), while the control point coordinates are introduced as random parameters (signals). A functional model relating the observations, parameters and signals is established, and the parameter estimates are subsequently obtained through fitting based on the LSC principle. Consequently, prior to constructing the error variance model, approximate values of the parameters—namely, the approximate coordinates of the station point—must first be computed. As shown in Fig. 7, the projection of point \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\angle\:SCF{III}_{i}CF{III}_{j}$$\end{document} onto the x–y plane is given by15:44\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\angle\:{S}^{{\prime\:}}CF{III}_{i}CF{{III}_{j}}^{{\prime\:}}={arccos}\frac{{D}_{i}^{2}+{D}_{ij}^{2}-{D}_{j}^{2}}{2{D}_{i}{D}_{ij}}$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${D}_{i}={S}_{i}\cdot\:{sin}{A}_{i}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${D}_{j}={S}_{j}\cdot\:{sin}{A}_{j}$$\end{document}, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${D}_{ij}=\sqrt{({X}_{j}-{X}_{i}{)}^{2}+({Y}_{j}-{Y}_{i}{)}^{2}}$$\end{document}. Fig. 7Schematic illustration of approximate station-point coordinate determination. Schematic illustration of approximate station-point coordinate determination. The azimuth angle from point \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$CF{III_{i}}$$\end{document} to point \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:CF{III}_{j}$$\end{document} is calculated as:45\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{T}_{i}}_{j}={arctan}\frac{{Y}_{j}-{Y}_{i}}{{X}_{j}-{X}_{i}}$$\end{document} Accordingly, the azimuth angle from the station point S to point \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:CF{III}_{i}$$\end{document} is given by:46\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{T}_{Si}={T}_{ij}-\angle\:{S}^{{\prime\:}}CFII{I}_{i}CFII{{I}_{j}}^{{\prime\:}}\pm\:180^\circ\:$$\end{document} Accordingly, the approximate 3D coordinates of the station point S are given by:47\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{l}{X}_{S}^{0}={X}_{i}-{S}_{i}{sin}{A}_{i}{cos}{T}_{i}\\\:{Y}_{S}^{0}={Y}_{i}-{S}_{i}{sin}{A}_{i}{sin}{T}_{i}\\\:{Z}_{s}^{0}={Z}_{i}-{S}_{i}{cos}{A}_{i}\end{array}\right.$$\end{document} Construction of the adjustment functional model Based on the 3D geometric relationship between the station point and the control points, the original observations—namely the distance \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{S}_{i}$$\end{document}, the horizontal direction \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{L}_{i}$$\end{document} and the zenith distance \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{A}_{i}$$\end{document}—their corresponding corrections \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{Si}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{Li}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{Ai}$$\end{document}, and the coordinates \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:({X}_{i},{Y}_{i},{Z}_{i})$$\end{document} of the control points \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:CF{III}_{i}$$\end{document} as well as the station-point coordinates \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:({X}_{S},{Y}_{S},{Z}_{S})$$\end{document} are related by the following expressions:48\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{S}_{i}+{v}_{Si}=\sqrt{({X}_{i}-{X}_{S}{)}^{2}+({Y}_{i}-{Y}_{S}{)}^{2}+({Z}_{i}-{Z}_{S}{)}^{2}}$$\end{document}49\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{tan}({L}_{i}+{v}_{Li}+\omega\:)=\frac{{Y}_{i}-{Y}_{S}}{{X}_{i}-{X}_{S}}$$\end{document}50\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{cos}({A}_{i}+{v}_{Ai})=\frac{{Z}_{i}-{Z}_{S}}{\sqrt{({X}_{i}-{X}_{S}{)}^{2}+({Y}_{i}-{Y}_{S}{)}^{2}+({Z}_{i}-{Z}_{S}{)}^{2}}}$$\end{document} The station-point coordinates are expressed as the sum of their approximate values \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:({X}_{S}^{0},{Y}_{S}^{0},\:{Z}_{S}^{0})$$\end{document} and the corresponding corrections \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:(d{X}_{S},d{Y}_{S},d{Z}_{S})$$\end{document}. Meanwhile, the control point coordinates are introduced as random parameters with known prior precision; accordingly, the control point coordinates are represented by the estimates \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:({X}_{i}^{0},{Y}_{i}^{0},{Z}_{i}^{0})$$\end{document} of their true values \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:({X}_{i},{Y}_{i},{Z}_{i})$$\end{document}, together with the corresponding corrections \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:(d{X}_{i},d{Y}_{i},d{Z}_{i})$$\end{document}. Error equation for distance observations By performing a Taylor series expansion of Eq. (48) and neglecting second- and higher-order terms, the error equation for the slope distance can be expressed as:51\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{Si}=-\frac{{X}_{i}^{0}-{X}_{S}^{0}}{{S}_{i}^{0}}\cdot\:d{X}_{S}-\frac{{Y}_{i}^{0}-{Y}_{S}^{0}}{{S}_{i}^{0}}\cdot\:d{Y}_{S}-\frac{{Z}_{i}^{0}-{Z}_{S}^{0}}{{S}_{i}^{0}}\cdot\:d{Z}_{S}+\frac{{X}_{i}^{0}-{X}_{S}^{0}}{{S}_{i}^{0}}\cdot\:d{X}_{i}+\frac{{Y}_{i}^{0}-{Y}_{S}^{0}}{{S}_{i}^{0}}\cdot\:d{Y}_{i}+\frac{{Z}_{i}^{0}-{Z}_{S}^{0}}{{S}_{i}^{0}}\cdot\:d{Z}_{i}-({S}_{i}-{S}_{i}^{0})$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{S}_{i}^{0}=\sqrt{({X}_{i}^{0}-{X}_{S}^{0}{)}^{2}+({Y}_{i}^{0}-{Y}_{S}^{0}{)}^{2}+({Z}_{i}^{0}-{Z}_{S}^{0}{)}^{2}}$$\end{document}, that is, the approximate value of the slope distance \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{S}_{i}$$\end{document}. Error equation for horizontal direction observations In Eq. (49), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\omega\:$$\end{document} denotes the unknown orientation angle of the station point, expressed as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\omega\:={\omega\:}^{0}+d\omega\:$$\end{document}, where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:d\omega\:$$\end{document} is the correction to the unknown orientation angle and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\omega\:}^{0}$$\end{document} is its approximate value. The calculation of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\omega\:}^{0}$$\end{document} is given as follows:52\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\omega\:}^{0}=\frac{{\sum\:}_{i=1}^{n}({T}_{i}^{0}-{L}_{i})}{n}$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{T}_{i}^{0}$$\end{document} denotes the approximate azimuth of each backsight direction, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{T}_{i}^{0}={arctan}\frac{{Y}_{i}^{0}-{Y}_{S}^{0}}{{X}_{i}^{0}-{X}_{S}^{0}}$$\end{document}, n denotes the number of backsight control points. By performing a Taylor series expansion of Eq. (49) and neglecting second- and higher-order terms, the error equation for the horizontal direction observations can be derived as:53\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{Li}=\rho\:\frac{{Y}_{i}^{0}-{Y}_{S}^{0}}{({D}_{i}^{0}{)}^{2}}d{X}_{S}-\rho\:\frac{{X}_{i}^{0}-{X}_{S}^{0}}{({D}_{i}^{0}{)}^{2}}d{Y}_{S}-\rho\:\frac{{Y}_{i}^{0}-{Y}_{S}^{0}}{({D}_{i}^{0}{)}^{2}}d{X}_{i}+\rho\:\frac{{X}_{i}^{0}-{X}_{S}^{0}}{({D}_{i}^{0}{)}^{2}}d{Y}_{i}-d\omega\:-({L}_{i}+{\omega\:}^{0}-{T}_{i}^{0})$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\rho\:={\frac{\left(180\times\:3600\right)}{\pi\:}}^{{\prime\:}{\prime\:}}\approx\:206264.80{6}^{{\prime\:\prime\:}}$$\end{document}, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{i}^{0}$$\end{document} denotes the approximate horizontal distance, expressed as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{i}^{0}=\sqrt{({X}_{i}^{0}-{X}_{S}^{0}{)}^{2}+({Y}_{i}^{0}-{Y}_{S}^{0}{)}^{2}}$$\end{document}. Error equation for zenith distance observations By performing a Taylor series expansion of Eq. (50) and neglecting second- and higher-order terms, the error equation for the zenith distance can be derived as:54\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}v_{Ai} = &-\rho\frac{(X_i^0-X_S^0)(Z_i^0-Z_S^0)}{(S_i^0)^2D_i^0}dX_S-\rho\frac{(Y_i^0-Y_S^0)(Z_i^0-Z_S^0)}{(S_i^0)^2D_i^0}dY_S+\rho\frac{D_i^0}{(S_i^0)^2}dZ_S \\ &+\rho\frac{(X_i^0-X_S^0)(Z_i^0-Z_S^0)}{(S_i^0)^2D_i^0}dX_i+\rho\frac{(Y_i^0-Y_S^0)(Z_i^0-Z_S^0)}{(S_i^0)^2D_i^0}dY_i-\rho\frac{D_i^0}{(S_i^0)^2}dZ_i-(A_i-A_i^0)\end{aligned}$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{A}_{i}^{0}$$\end{document} denotes the approximate zenith distance, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{A}_{i}}^{0}={arccos}\frac{{{Z}_{i}}^{0}-{{Z}_{S}}^{0}}{\sqrt{({{X}_{i}}^{0}-{{X}_{S}}^{0}{)}^{2}+({{Y}_{i}}^{0}-{{Y}_{S}}^{0}{)}^{2}+({{Z}_{i}}^{0}-{{Z}_{S}}^{0}{)}^{2}}}$$\end{document}. Adjustment computation and accuracy assessment By rearranging the error equations of the three types of observations according to Eq. (43), let:55\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:dX={\left[\begin{array}{ccccccc}d{X}_{1}&\:d{Y}_{1}&\:d{Z}_{1}&\:\cdots\:&\:d{X}_{n}&\:d{Y}_{n}&\:d{Z}_{n}\end{array}\right]}^{T}$$\end{document}56\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:dY={\left[\begin{array}{cccc}d{X}_{S}&\:d{Y}_{S}&\:d{Z}_{S}&\:d\omega\:\end{array}\right]}^{T}$$\end{document} Where n denotes the number of control points. By expressing the error equations in matrix form, the following representation is obtained:57\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:V={A}_{3n\times\:3n}\underset{3n\times\:1}{\left[\begin{array}{c}d{X}_{1}\\\:d{Y}_{1}\\\:d{Z}_{1}\\\:\vdots \\\:d{X}_{n}\\\:d{Y}_{n}\\\:d{Z}_{n}\end{array}\right]}+{B}_{3n\times\:4}\underset{4\times\:1}{\left[\begin{array}{c}d{X}_{S}\\\:d{Y}_{S}\\\:d{Z}_{S}\\\:d\omega\:\end{array}\right]}-\underset{3n\times\:1}{\left[\begin{array}{c}{S}_{1}-{S}_{1}^{0}\\\:{L}_{1}+{\omega\:}^{0}-{T}_{1}^{0}\\\:{A}_{i}-{A}_{i}^{0}\\\:\vdots \\\:{S}_{n}-{S}_{n}^{0}\\\:{L}_{n}+{\omega\:}^{0}-{T}_{n}^{0}\\\:{A}_{n}-{A}_{n}^{0}\end{array}\right]}$$\end{document} The signal coefficient matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{A}_{3n\times\:3n}$$\end{document} is given by:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left[\begin{array}{ccccccc}\frac{{X}_{1}^{0}-{X}_{S}^{0}}{{S}_{1}^{0}}&\:\frac{{Y}_{1}^{0}-{Y}_{S}^{0}}{{S}_{1}^{0}}&\:\frac{{Z}_{1}^{0}-{Z}_{S}^{0}}{{S}_{1}^{0}}&\:\cdots\:&\:0&\:0&\:0\\\:-\rho\:\frac{{Y}_{1}^{0}-{Y}_{S}^{0}}{({D}_{1}^{0}{)}^{2}}&\:\rho\:\frac{{X}_{1}^{0}-{X}_{S}^{0}}{({D}_{1}^{0}{)}^{2}}&\:0&\:\cdots\:&\:0&\:0&\:0\\\:\rho\:\frac{({X}_{1}^{0}-{X}_{S}^{0})({Z}_{1}^{0}-{Z}_{S}^{0})}{({S}_{1}^{0}{)}^{2}{\cdot\:D}_{1}^{0}}&\:\rho\:\frac{({Y}_{1}^{0}-{Y}_{S}^{0})({Z}_{1}^{0}-{Z}_{S}^{0})}{({S}_{1}^{0}{)}^{2}\cdot:{D}_{1}^{0}}&\:-\rho\:\frac{{D}_{1}^{0}}{({S}_{1}^{0}{)}^{2}}&\:\cdots\:&\:0&\:0&\:0\\\:\vdots &\:\vdots &\:\vdots &\:\ddots\:&\: \vdots &\: \vdots &\: \vdots \\\:0&\:0&\:0&\:\cdots\:&\:\frac{{X}_{n}^{0}-{X}_{s}^{0}}{{S}_{n}^{0}}&\:\frac{{X}_{n}^{0}-{X}_{s}^{0}}{{S}_{n}^{0}}&\:\frac{{Z}_{n}^{0}-{Z}_{s}^{0}}{{S}_{n}^{0}}\\\:0&\:0&\:0&\:\cdots\:&\:-\rho\:\frac{{Y}_{n}^{0}-{Y}_{s}^{0}}{({D}_{n}^{0}{)}^{2}}&\:\rho\:\frac{{X}_{n}^{0}-{X}_{s}^{0}}{({D}_{n}^{0}{)}^{2}}&\:0\\\:0&\:0&\:0&\:\cdots\:&\:\rho\:\frac{({X}_{n}^{0}-{X}_{s}^{0})({Z}_{n}^{0}-{Z}_{s}^{0})}{({S}_{n}^{0}{)}^{2}\cdot\:{D}_{n}^{0}}&\:\rho\:\frac{({Y}_{n}^{0}-{Y}_{s}^{0})({Z}_{n}^{0}-{Z}_{s}^{0})}{({S}_{n}^{0}{)}^{2}\cdot\:{D}_{n}^{0}}&\:-\rho\:\frac{{D}_{n}^{0}}{({S}_{n}^{0}{)}^{2}}\end{array}\right]$$\end{document} The parameter coefficient matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{B}_{3n\times\:4}$$\end{document} is given by:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left[\begin{array}{cccc}-\frac{{X}_{1}^{0}-{X}_{s}^{0}}{{S}_{1}^{0}}&\:-\frac{{Y}_{1}^{0}-{Y}_{s}^{0}}{{S}_{1}^{0}}&\:-\frac{{Z}_{1}^{0}-{Z}_{s}^{0}}{{S}_{1}^{0}}&\:0\\\:\rho\:\frac{{Y}_{1}^{0}-{Y}_{s}^{0}}{({D}_{1}^{0}{)}^{2}}&\:-\rho\:\frac{{X}_{1}^{0}-{X}_{s}^{0}}{({D}_{1}^{0}{)}^{2}}&\:0&\:-1\\\:-\rho\:\frac{({X}_{1}^{0}-{X}_{s}^{0})({Z}_{1}^{0}-{Z}_{s}^{0})}{({S}_{1}^{0}{)}^{2}\cdot\:{D}_{1}^{0}}&\:-\rho\:\frac{({Y}_{1}^{0}-{Y}_{s}^{0})({Z}_{1}^{0}-{Z}_{s}^{0})}{({S}_{1}^{0}{)}^{2}{\cdot\:D}_{1}^{0}}&\:\rho\:\frac{{D}_{1}^{0}}{({S}_{1}^{0}{)}^{2}}&\:0\\\:\vdots&\:\vdots &\: \vdots &\:\vdots \\\:-\frac{{X}_{n}^{0}-{X}_{s}^{0}}{{S}_{n}^{0}}&\:-\frac{{Y}_{n}^{0}-{Y}_{s}^{0}}{{S}_{n}^{0}}&\:-\frac{{Z}_{n}^{0}-{Z}_{s}^{0}}{{S}_{n}^{0}}&\:0\\\:\rho\:\frac{{Y}_{n}^{0}-{Y}_{s}^{0}}{({D}_{n}^{0}{)}^{2}}&\:-\rho\:\frac{{X}_{n}^{0}-{X}_{s}^{0}}{({D}_{n}^{0}{)}^{2}}&\:0&\:-1\\\:-\rho\:\frac{({X}_{n}^{0}-{X}_{s}^{0})({Z}_{n}^{0}-{Z}_{s}^{0})}{({S}_{n}^{0}{)}^{2}\cdot\:{D}_{n}^{0}}&\:-\rho\:\frac{({Y}_{n}^{0}-{Y}_{s}^{0})({Z}_{n}^{0}-{Z}_{s}^{0})}{({S}_{n}^{0}{)}^{2}\cdot\:{D}_{n}^{0}}&\:\rho\:\frac{{D}_{n}^{0}}{({S}_{n}^{0}{)}^{2}}&\:0\end{array}\right]$$\end{document} Based on the principle of LSC, the computation formula for the parameter corrections can be expressed as14,16:58\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:dY=\left[{B}^{T}\right(D+A{D}_{G}{A}^{T}{)}^{-1}B{]}^{-1}{B}^{T}(D+A{D}_{G}{A}^{T}{)}^{-1}L$$\end{document} Where D denotes the variance matrix of the observations, that is:59\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:D=diag\left[\begin{array}{cccc}{D}_{1}&\:{D}_{2}&\:\cdots\:&\:{D}_{n}\end{array}\right]$$\end{document}60\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{i}=diag\left[\begin{array}{ccc}(a+b\cdot\:{S}_{i}{)}^{2}&\:(\frac{{\sigma\:}_{\alpha\:}}{\rho\:}{)}^{2}&\:(\frac{{\sigma\:}_{A}}{\rho\:}{)}^{2}\end{array}\right]$$\end{document} Where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{G}$$\end{document} denotes the prior variance matrix of the signals, that is, the variance matrix of all control points:61\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{G}=diag\left[\begin{array}{cccc}{D}_{G1}&\:{D}_{G2}&\:\cdots\:&\:{D}_{Gn}\end{array}\right]$$\end{document}62\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{Gi}=diag\left[\begin{array}{ccc}{{\sigma\:}_{{x}_{i}}}^{2}&\:{{\sigma\:}_{{y}_{i}}}^{2}&\:{{\sigma\:}_{{z}_{i}}}^{2}\end{array}\right]$$\end{document} Accordingly, the adjusted parameter values—namely, the adjusted station-point coordinates and the orientation angle unknown—can be obtained using the following expression:63\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:Y={Y}^{0}+dY$$\end{document} Where Y denotes the parameter vector, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:Y={[\begin{array}{ccc}{X}_{S}&\:{Y}_{S}&\:\begin{array}{cc}{Z}_{S}&\:\omega\:\end{array}\end{array}\mathrm{]}}^{T}$$\end{document}; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{Y}^{0}$$\end{document} is the vector of approximate parameter values, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{Y}^{0}=[\begin{array}{ccc}{X}_{S}^{0}&\:{Y}_{S}^{0}&\:\begin{array}{cc}{Z}_{S}^{0}&\:{\omega\:}^{0}\end{array}\end{array}{]}^{T}$$\end{document}; and dY is the vector of parameter corrections, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:dY=[\begin{array}{ccc}d{X}_{S}&\:d{Y}_{S}&\:\begin{array}{cc}d{Z}_{S}&\:d\omega\:\end{array}\end{array}{]}^{T}$$\end{document}. According to the LSC principle10, the variance–covariance matrix of the parameter vector, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{Y}$$\end{document}, can be computed using the following expression:64\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{Y}={\left[{B}^{T}(D+A{D}_{G}{A}^{T}{)}^{-1}B\right]}^{-1}$$\end{document} By taking the square roots of the diagonal elements of the matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{Y}$$\end{document}, the standard deviations of the individual coordinate components (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{{x}_{S}}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{{y}_{S}}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{{z}_{S}}$$\end{document}) of the station point, as well as that of the orientation angle unknown (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{{\omega\:}_{S}}$$\end{document}), are obtained. Accordingly, the positional standard deviation of the station point S, denoted by \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{S}$$\end{document}, is given by:65\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{S}=\sqrt{{\sigma\:}_{{x}_{S}}^{2}+{\sigma\:}_{{y}_{S}}^{2}+{\sigma\:}_{{z}_{S}}^{2}}$$\end{document} Virtual simulation computation for 3D free-station positioning Simulation computation is a computer-model-based approach for accuracy assessment. By designing the network geometry and control point coordinates, the theoretical values of the observations are back-calculated and superimposed with random errors that conform to the statistical characteristics of real measurement errors17, thereby generating simulated datasets that are equivalent to actual observations in both accuracy and distribution18. Although free-station measurements eliminate errors associated with instrument centring and height measurement, the variability of station locations makes the geometric configuration and resulting accuracy difficult to evaluate in advance. Consequently, only coarse accuracy requirements can typically be specified at the design stage, while practical implementation often relies on on-site trial calculations and, if necessary, repeated re-stationing, which increases workload and entails a strong dependence on operator experience. Through numerical simulation of field observation processes, simulation computation enables a systematic evaluation of the accuracy of different free-station schemes during the engineering design phase. This allows the station region to be optimised and its expected accuracy to be estimated in advance, thereby effectively improving the efficiency and reliability of practical measurements. Principles and implementation of simulation computation Generation of pseudorandom numbers The core of virtual simulation computation lies in generating pseudorandom numbers that follow a normal distribution, in order to simulate the random errors encountered in field observations in accordance with statistical principles. First, uniformly distributed random numbers are generated and validated, after which the Box–Muller transformation is applied to obtain a sequence of standard normally distributed random variables19. The Box–Muller algorithm is based on a two-dimensional normal distribution, in which the two marginal distributions, X and Y, are mutually independent one-dimensional random variables that both follow a normal distribution20. Accordingly, the probability density function (PDF) of this two-dimensional normal distribution can be expressed as the product of the PDFs of two one-dimensional normal distributions:66\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:f(X,Y)=f\left(X\right)\cdot\:f\left(Y\right)=\frac{1}{\sqrt{2\pi\:}}{e}^{\frac{-{X}^{2}}{2}}\cdot\:\frac{1}{\sqrt{2\pi\:}}{e}^{\frac{-{Y}^{2}}{2}}=\frac{1}{2\pi\:}{e}^{\frac{-({X}^{2}+{Y}^{2})}{2}}=\frac{1}{2\pi\:}{e}^{\frac{-{r}^{2}}{2}}$$\end{document} The above expression indicates that this two-dimensional normal distribution can be described by a random vector r. In a polar coordinate system, r can be represented by two components—the angle θ and the magnitude r—that is:67\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{c}X=r\cdot\:{cos}\theta\:\\\:Y=r\cdot\:{sin}\theta\:\end{array}\right.$$\end{document} Assuming that X and Y both follow a standard normal distribution, integration of the above expression yields:68\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\int\:}_{-\infty\:}^{+\infty\:}{\int\:}_{-\infty\:}^{+\infty\:}\frac{1}{2\pi\:}{e}^{-\frac{{X}^{2}+{Y}^{2}}{2}}dXdY={\int\:}_{0}^{2\pi\:}{\int\:}_{0}^{+\infty\:}\frac{1}{2\pi\:}{e}^{-\frac{{r}^{2}}{2}}rdrd\theta\:=1$$\end{document} Accordingly, the distribution functions of r and θ can be obtained as:69\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left\{\begin{array}{c}{P}_{r}(r