Polynomial Chaos-based Input Shaper Design under Time-Varying Uncertainty

Johannes Güttler, Karan Baker, Premjit Saha, James Warner, Adrian Stein · 2026

Read Polynomial Chaos-based Input Shaper Design under Time-Varying Uncertainty by Johannes Güttler, Karan Baker, Premjit Saha, James Warner, Adrian Stein (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-17209.txt.

Introduction

Johannes Güttler, Karan Baker, Premjit Saha, James Warner, and Adrian Stein study how to keep machines steady when parts are not known for sure. They focus on input shapers. An input shaper times control signals so a system moves with less shake. Their tool is polynomial chaos expansion, a way to track doubt in math form.

The team puts special weight on doubt that changes over time. They use an intrusive form of the expansion, built right into the model. The goal is a shaper that stays robust as conditions drift. Such work helps engineers design controls that hold up in real use, not just on paper.

Catalog record

CREATORJohannes Güttler, Karan Baker, Premjit Saha, James Warner, Adrian Stein
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE28 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-17209

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Polynomial Chaos-based Input Shaper Design under Time-Varying Uncertainty Johannes Güttler, Karan Baker, Premjit Saha, James Warner, Adrian Stein Abstract The work presented here investigates the application of polynomial chaos expansion toward input shaper design in order to maintain robustness in dynamical systems subject to uncertainty. Furthermore, this work intends to specifically address time-varying uncertainty by employing intrusive polynomial chaos expansion. The methodology presented is validated through numerical simulation of intrusive polynomial chaos expansion formulation applied to spring mass system experiencing time-varying uncertainty in the spring stiffness. The system also evaluates non-robust and robust input shapers through the framework in order to identify designs that minimize residual energy. Results indicate that vibration mitigation is achieved at a similar accuracy, yet at higher efficiency compared to a Monte Carlo framework. I INTRODUCTION Uncertainty quantification (UQ) is vital for designing reliable and robust engineering systems, particularly when system parameters are subject to bounded intervals or evolve dynamically over time [ 18 , 20 , 21 , 24 , 23 ] . Classical probabilistic approaches often assume fixed distributions and time-invariant uncertainties, limiting their applicability for complex real-world scenarios such as Unmanned Aerial Vehicle-payload control, spacecraft attitude tracking, and structural dynamics [ 5 , 3 , 16 ] . Polynomial Chaos Expansion (PCE) provides an efficient spectral framework to represent stochastic system responses via orthogonal polynomial basis functions, enabling fast uncertainty propagation and sensitivity analysis [ 15 , 12 , 8 ] . Non-intrusive PCE approaches are widely used, but intrusive formulations, which embed uncertainty expansions directly into system equations, offer improved accuracy and computational benefits for dynamical systems [ 2 , 8 ] . Arbitrary PCEs have been developed, requiring only finite moments or raw data and allowing flexible data-driven UQ [ 6 ] . Hybrid methods combining PCE with set-based techniques such as zonotopic and interval analysis reduce conservatism and enhance interval estimation accuracy, particularly for nonlinear and fuzzy systems with time-invariant parameters [ 20 , 19 , 18 ] . Further, second-order uncertainty models and scalable probabilistic frameworks using, for instance, Johnson distributions enable unified treatment of aleatory and epistemic uncertainties without relying on independence or exact distributional forms [ 17 , 25 ] . Time-dependent interval process models and dynamic response analyses have also been introduced to capture evolving uncertainties in structures and control systems, where sliding mode control and iterative optimal attitude control methods can be applied for vibration control [ 21 , 23 , 24 ] . Despite these advances, it has been shown that classical PCE methods degrade in long-time integrations because the fixed orthogonal polynomial basis becomes suboptimal as the probability density functions (pdf) of the system response evolve, leading to a loss of accuracy and convergence; this motivates the development of adaptive or time-dependent PCE approaches. Time-dependent generalized polynomial chaos methods can adapt polynomial bases over time to maintain accuracy [ 1 ] . Surrogate models using sparse PCE combined with stochastic time warping have also improved oscillatory system analyses in the past [ 4 ] . In recent years, robust input shaping and sensitivity analyses that leverage global sensitivity metrics can facilitate uncertainty-aware control design in vibration control [ 10 , 12 , 13 , 9 ] . Time-delay filters (TDF) are the implementation framework of input shapers, where the shaper’s sequence of impulses is realized as delayed and weighted inputs to cancel system vibrations. Experimental applications include real-time input shaping for gantry cranes and minimum-time input shaping under velocity constraints [ 14 , 11 ] . Moreover, adaptive truncated PCE algorithms have been proposed for robust trajectory optimization in ascent launch vehicles facing nonlinear, fast time-varying uncertainties [ 16 ] . Building upon these foundations, this study proposes an intrusive PCE framework with time-dependent uncertainty in dynamical systems. The example showcased here is an undamped spring-mass system with uncertain spring stiffness over time. The novelty of this work is to design precision motion controllers under influence of time-varying uncertainties with global sensitivity analysis (GSA). By adapting the polynomial basis dynamically at specific time instants, this method enhances vibration suppression and robust motion control. The approach combines theoretical advances in PCE with practical applications to real-world systems. The study is organized as follows. Section II details the intrusive time-dependent PCE methodology. Section III presents numerical validations and comparisons between PCE and Monte Carlo (MC) simulations. Section IV concludes and discusses future research directions. II METHODOLOGY This section describes the PCE and TDF design strategies. Furthermore, the spring-mass system with its uncertainty is explained. Fig 1 illustrates a single spring-mass system with an input u . Fig. 1 : Spring-mass system with input u . The equation of motion of the spring-mass system can be written as: \displaystyle m\ddot{x}+kx=ku (1a) \displaystyle\iff\ddot{x}+\omega_{n}^{2}x=\omega_{n}^{2}u, (1b) where x is the position of the mass m . In this work it is assumed that the spring stiffness k is uncertain, which directly correlates to an uncertainty in the natural frequency by \omega_{n}=\sqrt{k/m} . The mass is assumed to be constant with m=1 . The residual energy for this system is defined as: \displaystyle V_{res}=\frac{m}{2}\dot{x}^{2}+\frac{k}{2}(x_{f}-x)^{2}, (2) where x_{f} is the desired position of the mass. II-A Polynomial Chaos Expansion Polynomial Chaos is a well-established approach for representing a stochastic output as a weighted combination of polynomial functions of the random variables as inputs. A PCE has the form of [ 8 ] : \displaystyle F(t,\zeta)=\sum^{\infty}_{i=0}a_{i}(t)\Psi_{i}(\zeta), (3) where F is the output variable, a_{i} are the PC-coefficients and \Psi_{i} are the basis polynomial functions. Moreover, t is the deterministic independent variable and \zeta is the random variable. The orthogonality condition of the basis functions is given by [ 18 ] : \displaystyle\left<\Psi_{i}(\zeta),\Psi_{j}(\zeta)\right>=\int_{\Omega}\Psi_{i}(\zeta)\Psi_{j}(\zeta)f(\zeta)d\zeta=g_{i}^{2}\delta_{ij}, (4) where g_{i}^{2} are the coefficients resulting from the evaluation of the inner product, and \delta_{ij} is the Kronecker delta. f(\zeta) is the pdf of \zeta . The Wiener–Askey scheme [ 22 ] provides the appropriate choice of polynomial basis functions for random variables with standard well-known distributions, as illustrated in Table I . TABLE I : Wiener-Askey scheme for polynomial chaos expansion. k_{l} and k_{u} are the user-defined lower and upper bounds. Distribution of \zeta Polynomial family Support Gaussian Hermite \left(-\infty,\infty\right) gamma Laguerre \left[0,\infty\right) beta Jacobi \left[k_{l},k_{u}\right] uniform Legendre \left[k_{l},k_{u}\right] The PC-coefficients can be determined using an intrusive or a non-intrusive approach. The intrusive approach requires the evaluation of inner products and may not yield closed-form expressions for certain nonlinear functions [ 7 ] . The non-intrusive approach can be considered a black-box approach, which requires commonly solving a least-squares problem to determine the coefficients [ 18 ] . In this work, the intrusive approach is used, which means that the coefficients are determined by exploiting Galerkin projections. Consider ( 3 ) which can be approximated as: \displaystyle F(t,\zeta)\approx F_{PC}=\sum^{P}_{i=0}a_{i}(t)\Psi_{i}(\zeta). (5) The left-hand side of ( 5 ) represents the known equation that need to be approximated by a series of P orthogonal polynomials with unknown coefficients a_{i} , as shown on the right-hand side of ( 5 ). Applying Galerkin’s method corresponds to taking the inner product of both sides of ( 5 ) with \Psi_{j}(\zeta) . In combination with the orthogonality condition, this results in: \displaystyle\int F(t,\zeta)\Psi_{j}(\zeta)f(\zeta)d\zeta=g_{j}^{2}a_{j}(t). (6) Eqn. ( 6 ) can be solved to determine

Read the complete text