From Stabilizing Regions to Certified Controllers: Closing the Selection Gap in Unified PID/PI Analysis for Time-Delay Plants
Şenol Gülgönül · 2026
Read From Stabilizing Regions to Certified Controllers: Closing the Selection Gap in Unified PID/PI Analysis for Time-Delay Plants by Şenol Gülgönül (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-2606-24202.txt.
Introduction
Gülgönül closes two gaps left by a recent unified method for tuning PID controllers on plants whose responses are delayed. The earlier framework maps regions of stabilising gains. But it leaves users to test interior points by hand to count unstable poles, and offers no rule for picking one controller. This 2026 note supplies both steps.
The pole count turns out to be available analytically. Through companion-matrix or Routh counts when no delay is present, and through an argument-principle evaluation for delay loops. Labelling every region with its count makes the whole partition decided, with no manual tests. Gülgönül then adds a time-domain selection rule. Among controllers giving monotone step responses, guaranteed by external positivity of the closed loop, choose the gains with minimum settling time. The result is one certified controller, not a map of maybes.
Catalog record
| CREATOR | Şenol Gülgönül |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 18 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.24202
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
From Stabilizing Regions to Certified Controllers: Closing the Selection Gap in Unified PID/PI Analysis for Time-Delay Plants Şenol Gülgönül Abstract A recent unified treatment of PID tuning for time-delay plants (An, Tang, Sun, Zhang and Chen, Automatica , 2026) combines D-partition with a boundary gradient vector (BGV) to orient the boundaries of stabilizing, relative-stability and stability-margin regions. That method answers a feasibility question, namely where admissible gains lie, and the authors concede that a manual interior-point test is still required to fix the unstable-pole count in each cell, and that selecting a single controller is left to the user. This note makes three contributions. First, we observe that the one operation the BGV leaves manual, the absolute unstable-pole count, is available analytically: exactly for delay-free designs through a companion-matrix or Routh count, and through an argument-principle (Mikhailov) evaluation for retarded-type delay loops. Labelling every cell with its analytic count removes the interior-point test and renders the partition fully decided. Second, we add the step the BGV framework cannot reach, a time-domain selection rule that returns one certified controller: among monotone step responses we choose the minimum-settling-time PI gains, characterized by a tangency condition, with monotonicity guaranteed by external positivity (a nonnegative closed-loop impulse response). Third, we flag a neutral-type pitfall that the unified analysis never delimits: an ideal PID with derivative action on a first-order-plus-dead-time (FOPTD) plant is neutral type, with a root chain on the imaginary axis when k K d = T kK_{d}=T . We reproduce the authors’ delay-free benchmark (Example 2) exactly, recovering both admissible K p K_{p} intervals, and demonstrate the complete pipeline on a FOPTD plant, delivering a certified monotone, fast-settling PI controller that the region-only method can neither locate nor justify. All claims are validated numerically. 1 Introduction PID and PI tuning for plants with dead time remains a live problem because the closed-loop characteristic equation is transcendental and the stabilizing set is awkward to describe [ 6 , 7 ] . The D-partition (D-decomposition) method maps the imaginary-axis crossings of the characteristic quasipolynomial into curves or surfaces that cut parameter space into cells, each with a fixed number of right-half-plane (RHP) poles [ 12 , 11 ] . The classical difficulty is that the partition alone does not say which cell is stable; one traditionally selects an interior point in every cell and counts RHP roots, which for a transcendental equation is expensive and must be repeated per cell. An, Tang, Sun, Zhang and Chen [ 1 ] address this with a boundary gradient vector (BGV), the gradient of the real part of the critical imaginary root with respect to the tuned parameters, evaluated on the boundary. Its sign against a crossing direction tells whether the RHP count rises or falls across that boundary, and a geometric reading (the BGV is the boundary normal, oriented by the sign of a Jacobian) lets one mark the favourable direction with little computation. The framework is presented uniformly for two or three of the four parameters (K_{p},K_{i},K_{d},\tau) and for nominal, relative-stability and gain/phase-margin boundaries. The BGV is an elegant repackaging of the crossing-direction or root-tendency idea [ 8 , 9 , 10 ] , but it answers only a feasibility question. Two limitations remain, and the authors state both. First, the absolute RHP count in any cell still needs a manual interior test; the BGV gives only relative changes. Second, the method returns a region, not a controller, and offers no statement about the time-domain response: two controllers in the same stable cell can have completely different step responses, and the gain/phase margins used as robustness surrogates do not constrain overshoot or settling time. This note closes both gaps and corrects one omission. 1. Analytic counting (Section 4 ). The interior-point test is unnecessary. For delay-free designs the count is the exact number of RHP eigenvalues of the companion matrix, equivalently a Routh test. For retarded-type delay loops it is an argument-principle (Mikhailov) evaluation along a contour enclosing the RHP. Either way every cell is labelled directly, and combined with the BGV crossing signs a single anchor decides the whole partition. 2. Certified selection (Section 5 ). We add the missing stage. Within the feasible set we select, by a tangency condition, the monotone minimum-settling-time PI controller [ 15 , 16 ] , and we certify monotonicity by external positivity, a nonnegative closed-loop impulse response. This returns one controller with a guarantee on its response. 3. A neutral-type pitfall (Section 3 ). The unified analysis does not delimit its plant class. An ideal PID on a FOPTD plant is neutral type; at kK_{d}=T a chain of roots hugs the imaginary axis. We exhibit it and note the implied constraint |kK_{d}|<T . Section 6 validates everything: Example 2 of [ 1 ] is reproduced exactly, and the full pipeline is demonstrated on a FOPTD plant. Section 7 states plainly what is improved, what merely re-derives existing results, and what remains open (the genuine four-parameter region). 2 Problem setup and the BGV method Consider a linear time-delay plant G_{p}(s)=\frac{N(s)}{D(s)}e^{-\tau s} controlled by a PID law C(s)=K_{p}+K_{i}/s+K_{d}s . The closed-loop characteristic quasipolynomial is F(s)=sD(s)+\bigl(K_{d}s^{2}+K_{p}s+K_{i}\bigr)N(s)\,e^{-\tau s}. (1) Writing F(j\omega)=F_{r}(\omega,q)+jF_{i}(\omega,q) with q the tuned parameters, the D-partition boundaries are F_{r}=F_{i}=0 . The BGV of [ 1 ] is, componentwise, \nabla_{k}=-\frac{1}{\Omega}\!\left(\frac{\partial F_{r}}{\partial q_{k}}\frac{\partial F_{i}}{\partial\omega}-\frac{\partial F_{r}}{\partial\omega}\frac{\partial F_{i}}{\partial q_{k}}\right),\qquad\Omega=\Bigl(\tfrac{\partial F_{r}}{\partial\omega}\Bigr)^{2}+\Bigl(\tfrac{\partial F_{i}}{\partial\omega}\Bigr)^{2}, (2) which is the root sensitivity \mathrm{Re}(\partial s/\partial q_{k}) along the boundary, equal under one-dimensional reduction to the root tendency of [ 8 ] . The change in the RHP count along a direction v is \Delta n=\sum_{l}\operatorname{sgn}\bigl(v\cdot g(\omega_{l},q)\bigr) over the crossing frequencies \omega_{l} . This is a relative statement; the absolute count in any one cell is not supplied, and [ 1 ] obtains it by evaluating an interior point. 3 A neutral-type pitfall for ideal PID on FOPTD Before improving the method we record a structural caveat. Take the FOPTD plant G_{p}(s)=\frac{k}{Ts+1}e^{-Ls} with an ideal PID. From ( 1 ) the two highest-order terms are Ts^{2} and kK_{d}s^{2}e^{-Ls} , both of order s^{2} . The principal term is therefore s^{2}\bigl(T+kK_{d}e^{-Ls}\bigr) and the loop is of neutral type. The associated difference operator is stable only if |kK_{d}|<T, (3) and at kK_{d}=T the equation T+kK_{d}e^{-Ls}=0 has the purely imaginary solutions s=\pm j\pi(2m+1)/L , a chain of roots on the imaginary axis. Numerically, with k=T=1 , L=0.5 and K_{d}=1 , an RHP-count sweep returns a large, mesh-dependent number of near-axis roots rather than a finite figure, exactly the signature of ( 3 ) being violated at the boundary. The unified analysis of [ 1 ] never states this restriction, so a derivative-acting design read off its surfaces can be marginal or ill-posed. The clean and practically relevant slice is therefore PI control (or a properly filtered derivative), to which we now restrict. 4 Stage 1: complete feasible set with an analytic root count 4.1 Delay-free designs: exact count When \tau=0 , ( 1 ) is an ordinary polynomial and the RHP count is the number of its roots with positive real part, obtained exactly from the companion-matrix eigenvalues or a Routh array. No interior search and no transcendental root-finding are involved; the count is closed-form algebra. The admissible projection onto any one parameter