Pith. sign in

REVIEW 1 major objections 5 minor 48 references

Robust Optimal Control of Arbitrarily Switched Systems: A Path-Complete Framework

T0 review · 1 major / 5 minor · reviewed 2026-08-02 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Graph-based Bellman inequalities over path-complete graphs produce certified worst-case cost bounds and policies for arbitrarily switched systems.

desk verdict Theorem 1 is sound and the reachability-graph min–max construction is a real extension, but the general-graph synthesis story is conditional on an initialization heuristic with no proven feasibility or convergence conditions. read the letter →

arxiv 2607.15055 v1 pith:E2TJAUKH submitted 2026-07-16 math.OC

Léa Ninite , Adrien Banse , Guillaume O. Berger , Raphaël M. Jungers This is my paper
classification math.OC MSC 93C3090C3949N3593D09
keywords switchedsystemsarbitraryswitchingrobustcontrolpath-completegraphsBellmaninequalitiesvaluefunctionsemidefiniteprogrammingmodelpredictive
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that for a switched system whose mode is chosen adversarially at every step, controller design and performance certification can be done in one step by distributing the robust Bellman inequality over a path-complete graph. Instead of solving the intractable single Bellman equation, one enforces local graph-based Bellman inequalities on node-indexed functions and combines them pointwise via a min-max rule; the resulting function is a guaranteed upper bound on the closed-loop value function of the extracted policy. The practical promise is that this bound is not just a certificate: it is optimized during synthesis, so the controller is aimed at low worst-case cost from the start. For linear systems with quadratic costs the inequalities become matrix inequalities, solvable by a single semidefinite program for complete graphs and by alternating optimization in general. A building temperature benchmark shows the certificates can also be used as terminal costs in robust MPC, cutting online computation time by orders of magnitude compared with scenario-tree MPC while matching closed-loop cost.

What carries the argument

The load-bearing construction is a path-complete graph G=(S,E), a directed labeled graph in which every finite mode sequence is realized as a path, together with one of its reachability graphs H=(S_H,E_H), whose nodes are subsets of S and which is complete for every mode. Local quadratic functions V_alpha and linear policies pi_A are required to satisfy graph-based Bellman inequalities along edges; the reachability property ensures that whatever node set A is active, every mode i has an outgoing edge to some B whose every element is reachable from some element of A. The min-max combination V(x)=min_A max_{alpha in A} V_alpha(x), with policy pi_{A*(x)}, turns these local inequalities into the

What would settle it

On a small two-mode switched linear system with quadratic cost, compute the certificate from equation (5) by solving the matrix inequalities (13) for a complete path-complete graph of order 2, then simulate the resulting piecewise-linear policy under all mode sequences up to 20 steps from a fixed x0. If the finite-horizon empirical worst-case cost ever exceeds V(x0), the graph-based Bellman inequalities cannot have been satisfied and the certification claim would be refuted.

Watch

Extended reading notes

Core claim

The paper proves that if functions on the nodes of a path-complete graph and policies on the nodes of a reachability graph satisfy graph-based Bellman inequalities, then the pointwise min-max function V(x)=min_A max_{alpha in A} V_alpha(x), together with the policy phi(x)=pi_{A*(x)}(x), satisfies the robust Bellman inequality. Consequently V is a certified upper bound on the closed-loop value function of phi and hence also on the optimal value function J*. In the LQR setting these conditions become matrix inequalities; complete graphs give a single semidefinite program, while general path-complete graphs are handled by alternating optimization. Numerical results on a building temperature reg

Load-bearing premise

The practical pipeline assumes an initial stabilizing controller with a contraction factor below one can be found; for switched systems where no such stabilizing controller exists, the alternating optimization cannot be initialized.

Editorial extensions

If this is right

  • If the graph-based inequalities are feasible, the theorem delivers both a policy and a certified upper bound on its worst-case cost, so no separate verification step is needed.
  • For complete graphs the whole synthesis reduces to one semidefinite program; for general path-complete graphs, alternating optimization over the matrices P and the feedback gains K provides a tractable route.
  • Because the certified function satisfies the robust Bellman inequality, it can be appended as a terminal cost in robust MPC; in the reported benchmark this lowered the horizon-5 MPC average cost from about 1487 to 1197.
  • Co-complete graph structures can outperform complete ones on the same benchmark, showing that moving beyond complete graphs is practically relevant, not just a theoretical generalization.
  • The construction recovers earlier complete- and co-complete-graph results as special cases and extends them to controlled switched systems with inputs.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Extension: the gap between the certified V(x0) and the best empirical closed-loop cost can be read as a measure of how much information the graph's memory encodes about adversarial mode sequences; larger graphs should shrink it, and the paper's future-work note points in the same direction.
  • Extension: since V is pointwise evaluable and non-differentiable only at switching boundaries, it could serve as a candidate for safety-barrier or control-Lyapunov reuse, not just as a cost certificate.
  • Extension: the same graph-based relaxation idea should transfer to continuous-state robust Markov decision processes, replacing the expectation operator with the robust worst-case operator; a natural test is a continuous-state inventory or queueing problem with adversarial demand.
  • Extension: a practical refinement is to warm-start Algorithm 1 with a certificate from a coarser graph and then progressively refine the graph structure, which could cut the offline synthesis times reported at higher graph orders.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

1 major / 5 minor

Summary. The paper develops a path-complete graph framework for joint synthesis of feedback policies and certified upper bounds on the infinite-horizon worst-case value function for arbitrarily switched systems. Functions attached to graph nodes and policies attached to nodes of a reachability graph are constrained by graph-based Bellman inequalities (3). The main theorem (Theorem 1) asserts that the pointwise min-max combination V(x)=min_A max_{α∈A} V_α(x) is a certified upper bound on the closed-loop value function of the policy φ(x)=π_{A*(x)}(x), and hence on the optimal value function. Complete and co-complete graphs are recovered as corollaries. For linear switched systems with quadratic stage costs, the inequalities become matrix inequalities; complete graphs yield a single SDP (17)-(18), while general path-complete graphs are treated by alternating optimization (Algorithm 1) initialized by a stabilization heuristic (Appendix A). Numerical experiments compare the proposed controllers with robust MPC on a synthetic example and a three-zone building temperature regulation benchmark, including the use of the certified bound as an MPC terminal cost.

Significance. The theoretical result is significant and, as far as I can verify, correct. Theorem 1 provides a genuine extension of path-complete Lyapunov methods to certified performance bounds with explicit policy synthesis, and the reachability-graph construction is elegant. The LMI formulation for complete graphs is useful, and the terminal-cost application to robust MPC is a nice demonstration. The paper also releases code for reproduction. The main weakness is computational: Algorithm 1's initialization is a heuristic without proven feasibility or convergence, so the claim that general path-complete graphs are 'handled' is conditional. This is fixable and does not undermine the core theory.

major comments (1)
  1. [§5.1/Appendix A (Algorithm 1, line 2; Eq. (19))] Algorithm 1 requires an initial feasible solution to (13). Appendix A proposes to find one by alternating optimization on (19) seeking a contraction factor γ<1, but it gives no conditions under which such a solution exists, no guidance for choosing the starting positive definite matrices, and no convergence guarantee for the alternation. For a switched linear system that is not switched-stabilizable, or for a graph G that admits no stabilizing assignment of the prescribed form, no such γ<1 exists and the procedure cannot initialize; for stabilizable systems the heuristic can still fail on a given graph. Thus the paper does not delimit the class of systems/graphs for which the general path-complete synthesis in Algorithm 1 is guaranteed to run. The authors should either prove sufficient conditions for the existence of the initial feasible point (e.g., existence of a stabilizing path-compl
minor comments (5)
  1. [Definition 5 / Theorem 1] The nodes of a reachability graph are subsets of S, but nonemptiness is not stated. If an empty set were allowed, the expressions max_{α∈A}V_α(x), argmin, and the proof would be undefined. Please add an explicit nonemptiness requirement for all A∈S_H.
  2. [Proof of Theorem 1] After applying (3) to (α*, β*, i), the proof jumps from V(x) ≥ c(x,π_A(x)) + max_{β∈B}V_β(f_i(x,π_A(x))) to V(x) ≥ c(x,φ(x)) + V(f_i(x,φ(x))). The missing intermediate step is V(f_i(x,φ(x))) ≤ max_{β∈B}V_β(f_i(x,φ(x))), which follows from (5) because B is one of the sets in the min. Adding this sentence would make the proof fully explicit.
  3. [Section 6, Table 1] The empirical costs J^avg_T are averages over 50 Monte Carlo runs, but no standard deviations, confidence intervals, or per-trial spread are reported. Since the qualitative claims (e.g., 'nearly identical costs', 'improved performance') rely on these averages, please add error bars or a variance measure.
  4. [Remark 1] The statement that any graph admitting a reachability graph is path-complete is asserted with the proof omitted. This is not used later, so it is not load-bearing, but in a journal version the proof should be supplied or a reference given.
  5. [Section 6, building benchmark] The building model matrices are delegated to [44] rather than stated. For reproducibility, include the numerical A_i,B_i matrices (or an appendix) or make clear that they are fully specified in the accompanying code repository.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: Theorem 1's certification argument is self-contained; only minor non-load-bearing self-citations to the authors' prior work appear.

full rationale

The central derivation is not circular. Theorem 1 assumes graph-based Bellman inequalities (3) and constructs V and phi from the same V_alpha and pi_A; the proof then shows, using only the completeness of the reachability graph and condition (ii) of Definition 5, that (2) holds at every x. Proposition 2, which converts (2) into the bound V >= J^phi, is proved in the paper by a telescoping-sum argument and does not presuppose the theorem's conclusion. No parameter is fitted to observed costs or to J^phi; the synthesis minimizes trace(P) or gamma on the bound itself, and the reported V(x0) is a certified upper bound, not a prediction fitted to data. Self-citations to [33] occur only for Proposition 2 (whose proof is included) and for the standard Schur-complement LMI reformulation (17), and [3,17] are cited as prior constructions rather than as the load-bearing justification of Theorem 1. No uniqueness theorem is imported from the authors' own work, and no ansatz is smuggled via citation: the quadratic template and piecewise-linear policies are explicit modeling choices. The initialization feasibility concern in Appendix A is a real algorithmic gap but is a correctness/completeness issue, not a circularity: it does not make the certified bound equivalent to its inputs. Overall, the paper's main claim has independent content and is not forced by definition or by a self-citation chain.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The central theorem is self-contained given the problem definitions. The only non-standard assumption is the existence of a feasible initialization for the heuristic alternating algorithm. No physical entities are invented; the reachability graph is a mathematical construction, not an empirical postulate.

assumptions (4)
  • domain assumption The stage cost c is nonnegative (c: R^n x R^m -> R_{\ge0})
    Used throughout; ensures V(x_H) >= 0 in the telescoping proof of Proposition 2.
  • domain assumption The switching signal is fully adversarial: value function is the supremum over all mode sequences
    Defines the problem; robust Bellman equation relies on worst-case max over modes.
  • standard math Schur complement equivalences for LMI reformulations
    Used to convert PMI (13) to LMI (17) and (15) to SDP in Appendix B.
  • domain assumption Existence of an initial feasible solution to (13) in Algorithm 1
    Needed to initialize alternating optimization; not guaranteed for non-stabilizable systems.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Robust Optimal Control of Arbitrarily Switched Systems: A Path-Complete Framework." pith.science (2026). https://pith.science/paper/E2TJAUKH

@misc{pith2026260715055,
  author       = {Pith},
  title        = {Pith review of: Robust Optimal Control of Arbitrarily Switched Systems: A Path-Complete Framework},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/E2TJAUKH}},
  note         = {Machine review of arXiv:2607.15055}
}
read the original abstract

This paper addresses the robust control of switched systems under arbitrary switching with performance guarantees. We propose a framework that jointly synthesizes a feedback policy and a certified upper bound on its corresponding infinite-horizon closed-loop value function. The proposed upper bound not only certifies the performance of the synthesized policy, but can also be optimized during controller synthesis. More precisely, our approach associates functions with the nodes of a path-complete graph and enforces graph-based Bellman inequalities along its edges. Exploiting a newly introduced notion of reachability graph, these functions are combined into both a feedback policy and a certified upper bound on its corresponding closed-loop value function, expressed as a pointwise min-max combination of the graph-indexed functions. For linear switched systems with quadratic stage costs, the proposed framework admits tractable computational formulations based on semidefinite programming and alternating optimization. Numerical experiments, including a building temperature regulation benchmark, demonstrate the practical usefulness of the proposed approach both for direct feedback control using the synthesized policy and for model predictive control using the certified upper bound as a terminal cost.

Figures

Figures reproduced from arXiv: 2607.15055 by the authors.

Figure 1
Figure 1. Overview of the proposed approach. A path-complete graph is used to generate a collection of [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. (a) Path-complete graph with two nodes, for a system with two switching modes. (b) Graph not [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Path-complete graph with four nodes, for a system with two switching modes. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Two examples of reachability graphs for the graph in Figure [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Certified upper bounds V ℓ c and V ℓ cc on the value function, obtained from primal and dual De Bruijn graphs of order ℓ ∈ {1, 2, 3, 4}, respectively. The bounds V ℓ c are computed via the SDP (18), whereas V ℓ cc are computed using Algorithm 1. The bounds are evaluate…
Figure 6
Figure 6. Figure 6: Three-zone building temperature regulation benchmark. [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]
Figure 7
Figure 7. Figure 7: Closed-loop state trajectories under four controllers: the proposed path-complete controller based [PITH_FULL_IMAGE:figures/full_fig_p017_7.png]
Figure 8
Figure 8. Figure 8: Primal De Bruijn graphs of orders ℓ = 1, ℓ = 2, and ℓ = 3 for the mode set ⟨M⟩ = {1, 2}. Blue and orange edges correspond to labels 1 and 2, respectively. 22 [PITH_FULL_IMAGE:figures/full_fig_p022_8.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

48 extracted references · 2 linked inside Pith

  1. [44]

    Z. Wang, G. O. Berger, and R. M. Jungers. Data-driven control of unknown switched linear systems using scenario optimization.IEEE Transactions on Automatic Control, 69(11):7310–7325, 2024

    2024

  2. [33]

    Ninite, A

    L. Ninite, A. Banse, G. O. Berger, and R. M. Jungers. A path-complete approach for optimal control of switched systems. InProceedings of the 29th International Conference on Hybrid Systems: Computation and Control, 2026

    2026

  3. [1]

    Jointspectralradiusandpath-complete graph lyapunov functions.SIAM Journal on Control and Optimization, 52(1):687–717, 2014

    A.A.Ahmadi, R.M.Jungers, P.A.Parrilo, andM.Roozbehani. Jointspectralradiusandpath-complete graph lyapunov functions.SIAM Journal on Control and Optimization, 52(1):687–717, 2014

    2014

  4. [2]

    M. Alamir. Learning against uncertainty in control engineering.Annual Reviews in Control, 53:19–29, 2022

    2022

  5. [3]

    Angeli, N

    D. Angeli, N. Athanasopoulos, R. M. Jungers, and M. Philippe. Path-complete graphs and common lya- punov functions. InProceedings of the 20th International Conference on Hybrid Systems: Computation and Control, HSCC ’17, page 81–90, New York, NY, USA, 2017. Association for Computing Machinery

    2017

  6. [4]

    Baier and J.-P

    C. Baier and J.-P. Katoen.Principles Of Model Checking. MIT Press, 2008

    2008

  7. [5]

    Baier, P

    R. Baier, P. Braun, L. Grüne, and C. M. Kellett. Numerical calculation of nonsmooth control lyapunov functions via piecewise affine approximation∗∗p. braun, l. grüne and c. m. kellett are supported by the australian research council (grant number: Arc-dp160102138).IFAC-PapersOnLine, 52(16):370–375,

  8. [6]

    Belta, B

    C. Belta, B. Yordanov, and E. A. Gol.Formal Methods for Discrete-Time Dynamical Systems, volume 89 ofStudies in Systems, Decision and Control. Springer

Show all 48 references
  1. [7]

    Bemporad and M

    A. Bemporad and M. Morari. Robust model predictive control: A survey. In A. Garulli and A. Tesi, editors,Robustness in identification and control, pages 207–226, London, 1999. Springer London

    1999

  2. [8]

    G. O. Berger. On the differentiability of the value function of switched linear systems under arbitrary and controlled switching.arXiv preprint arXiv:2511.20037, 2025

    2025 arXiv

  3. [9]

    Bertsekas.Dynamic programming and optimal control: Volume I, volume 4

    D. Bertsekas.Dynamic programming and optimal control: Volume I, volume 4. Athena scientific, 2012

    2012

  4. [10]

    Blanchini

    F. Blanchini. Nonquadratic lyapunov functions for robust control.Automatica, 31(3):451–461, 1995

    1995

  5. [11]

    Blanchini, D

    F. Blanchini, D. Casagrande, P. Gardonio, and S. Miani. Constant and switching gains in semi-active damping of vibrating structures.International Journal of Control, 85(12):1886–1897, 2012

    2012

  6. [12]

    Blanchini, S

    F. Blanchini, S. Miani, et al.Set-theoretic methods in control, volume 78. Springer, 2008

    2008

  7. [13]

    Calbert, B

    J. Calbert, B. Legat, L. N. Egidio, and R. Jungers. Alternating simulation on hierarchical abstractions. In2021 60th IEEE Conference on Decision and Control (CDC), pages 593–598, 2021

    2021

  8. [14]

    Chang, N

    Y.-C. Chang, N. Roohi, and S. Gao. Neural lyapunov control. In H. Wallach, H. Larochelle, A. Beygelz- imer, F. d'Alché-Buc, E. Fox, and R. Garnett, editors,Advances in Neural Information Processing Systems, volume 32. Curran Associates, Inc., 2019

    2019

  9. [15]

    N. G. De Bruijn and P. Erdös. On a combinatorial problem.Proceedings of the Section of Sciences of the Koninklijke Nederlandse Akademie van Wetenschappen te Amsterdam, 51(10):1277–1279, 1948

    1948

  10. [16]

    de Oliveira, J

    M. de Oliveira, J. Bernussou, and J. Geromel. A new discrete-time robust stability condition.Systems & Control Letters, 37(4):261–265, 1999

    1999

  11. [17]

    Della Rossa, T

    M. Della Rossa, T. Alves Lima, M. Jungers, and R. M. Jungers. Graph-based conditions for feedback stabilization of switched and lpv systems.Automatica, 160:111427, 2024

    2024

  12. [18]

    Della Rossa and R

    M. Della Rossa and R. M. Jungers. Multiple lyapunov functions and memory: A symbolic dynamics approach to systems and control.SIAM Journal on Control and Optimization, 62(5):2695–2722, 2024

    2024

  13. [19]

    Donkers, W

    M. Donkers, W. M. H. Heemels, N. Van de Wouw, and L. Hetel. Stability analysis of networked control systems using a switched linear systems approach.IEEE Transactions on Automatic control, 56(9):2101–2115, 2011. 18

    2011

  14. [20]

    Fujita, T

    T. Fujita, T. Ueno, and S. Iwamoto. A nondeterministic dynamic programming model. InInternational Conference on Knowledge-Based and Intelligent Information and Engineering Systems, pages1208–1214. Springer, 2004

    2004

  15. [21]

    Görges, M

    D. Görges, M. Izák, and S. Liu. Optimal control and scheduling of switched systems.IEEE Transactions on Automatic Control, 56(1):135–140, 2011

    2011

  16. [22]

    E. A. Hernandez-Vargas, R. H. Middleton, and P. Colaneri. Optimal and mpc switching strategies for mitigating viral mutation and escape.IFAC Proceedings Volumes, 44(1):14857–14862, 2011

    2011

  17. [23]

    G. N. Iyengar. Robust dynamic programming.Mathematics of Operations Research, 30(2):257–280, 2005

    2005

  18. [24]

    Jungers.The joint spectral radius: theory and applications, volume 385

    R. Jungers.The joint spectral radius: theory and applications, volume 385. Springer Science & Business Media, 2009

    2009

  19. [25]

    E. A. Lee and S. A. Seshia.Introduction to Embedded Systems A Cyber-Physical Systems Approach. MIT Press, 2016

    2016

  20. [26]

    Liberzon.Switching in systems and control, volume 190

    D. Liberzon.Switching in systems and control, volume 190. Springer, 2003

    2003

  21. [27]

    T. A. Lima, M. Della Rossa, and A. Girard. Feedback stabilization of switched systems under arbitrary switching: A convex characterization.arXiv preprint arXiv:2506.03759, 2025

    2025 arXiv

  22. [28]

    Lincoln and A

    B. Lincoln and A. Rantzer. Relaxing dynamic programming.IEEE Transactions on Automatic Control, 51(8):1249–1260, 2006

    2006

  23. [29]

    F. Lu, P. G. Mehta, S. P. Meyn, and G. Neu. Convex q-learning. In2021 American Control Conference (ACC), pages 4749–4756. IEEE, 2021

    2021

  24. [30]

    V. L. Mehrmann.The autonomous linear quadratic control problem: theory and numerical solution. Springer, 1991

    1991

  25. [31]

    Mitra.Verifying Cyber-Physical Systems: A Path to Safe Autonomy

    S. Mitra.Verifying Cyber-Physical Systems: A Path to Safe Autonomy. MIT Press, 2021

    2021

  26. [32]

    Nilim and L

    A. Nilim and L. E. Ghaoui. Robust control of markov decision processes with uncertain transition matrices.Operations Research, 53(5):780–798, 2005

    2005

  27. [34]

    Pairet, J

    E. Pairet, J. D. Hernandez, M. Carreras, Y. Petillot, and M. Lahijanian. Online mapping and mo- tion planning under uncertainty for safe navigation in unknown environments.IEEE Transactions on Automation Science and Engineering, 19(4):3356–3378, Oct. 2022

    2022

  28. [35]

    P. A. Parrilo. Semidefinite programming relaxations for semialgebraic problems.Mathematical Pro- gramming, Series B, 96, 2003

    2003

  29. [36]

    Prajna and A

    S. Prajna and A. Jadbabaie. Safety verification of hybrid systems using barrier certificates. In R. Alur and G. J. Pappas, editors,Hybrid Systems: Computation and Control, pages 477–492, Berlin, Heidel- berg, 2004. Springer Berlin Heidelberg

    2004

  30. [37]

    Prajna, A

    S. Prajna, A. Jadbabaie, and G. J. Pappas. A framework for worst-case and stochastic safety verification using barrier certificates.IEEE Transactions on Automatic Control, 52(8):1415–1428, 2007

    2007

  31. [38]

    A. Rantzer. On approximate dynamic programming in switching systems. InProceedings of the 44th IEEE Conference on Decision and Control, pages 1391–1396. IEEE, 2005. 19

    2005

  32. [39]

    J. B. Rawlings, D. Q. Mayne, and M. M. Diehl.Model Predictive Control Theory, Computation, and Design, 2nd Edition. Nob Hill Publishing, LLC

  33. [40]

    C. A. Sanchez, G. Garcia, S. Hadjeras, W. M. H. Heemels, and L. Zaccarian. Practical stabiliza- tion of switched affine systems with dwell-time guarantees.IEEE Transactions on Automatic Control, 64(11):4811–4817, 2019

    2019

  34. [41]

    Suilen, T

    M. Suilen, T. Badings, E. M. Bovy, D. Parker, and N. Jansen. Robust markov decision processes: A place where ai and formal methods meet. InPrinciples of Verification: Cycling the Probabilistic Landscape: Essays Dedicated to Joost-Pieter Katoen on the Occasion of His 60th Birth...

    2024

  35. [42]

    Sun and S

    Z. Sun and S. S. Ge. Stability theory of switched dynamical systems. 2011

    2011

  36. [43]

    Tabuada.Verification and Control of Hybrid Systems: A Symbolic Approach

    P. Tabuada.Verification and Control of Hybrid Systems: A Symbolic Approach. Springer US, 2009

    2009

  37. [45]

    Wu and Q

    Z. Wu and Q. He. Optimal switching sequence for switched linear systems.SIAM Journal on Control and Optimization, 58(2):1183–1206, 2020

    2020

  38. [46]

    Zhang, J

    W. Zhang, J. Hu, and A. Abate. On the value functions of the discrete-time switched lqr problem. IEEE Transactions on Automatic Control, 54(11):2669–2674, 2009

    2009

  39. [47]

    Zhang, J

    W. Zhang, J. Hu, and A. Abate. Infinite-horizon switched lqr problems in discrete time: A suboptimal algorithm with performance analysis.IEEE Transactions on Automatic Control, 57(7):1815–1821, 2012. A Initialization of Algorithm 1 The alternating optimization procedure descri...

    2012

  40. [2019]

    11th IFAC Symposium on Nonlinear Control Systems NOLCOS 2019

    2019

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.