arxiv: 2604.03415 · v1 · submitted 2026-04-03 · 📡 eess.SY · cs.SY· math.OC

Two-Timescale Asymptotic Simulations of Hybrid Inclusions with Applications to Stochastic Hybrid Optimization

Max F. Crisafulli , Andrew R. Teel This is my paper

Pith reviewed 2026-05-13 18:54 UTC · model grok-4.3

classification 📡 eess.SY cs.SYmath.OC

keywords hybrid inclusionstwo-timescale simulationsingular perturbationstochastic approximationhybrid optimizationweakly invariant setsboundary layer systemreduced system

0 comments

The pith

Sufficient conditions are given under which sequences of iterates and step sizes form a two-timescale asymptotic simulation of singularly perturbed hybrid inclusions, with limits characterized by weakly invariant sets of boundary layer and

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

The paper develops convergence results for model-free two-timescale simulations of hybrid inclusions that combine differential and difference inclusions to model flow and jump dynamics. It gives sufficient conditions on iterates and step sizes so that the sequences behave like a singularly perturbed system whose limiting behavior is captured by the weakly invariant and internally chain-transitive sets of an associated boundary layer system and reduced system. The same conditions are applied to show that a two-timescale stochastic approximation of a hybrid optimization algorithm asymptotically recovers the trajectories of its deterministic counterpart. A reader would care because the results supply tools for analyzing convergence in systems that mix continuous evolution with discrete resets, especially when stochastic noise is present.

Core claim

The central claim is that for singularly perturbed hybrid inclusions, under stated conditions on the sequences of iterates and step sizes, the sequences constitute a two-timescale asymptotic simulation whose limiting behavior is characterized via weakly invariant and internally chain-transitive sets of an associated boundary layer system and reduced system. This framework is used to prove that a two-timescale stochastic approximation of a hybrid optimization algorithm asymptotically recovers the behavior of the corresponding deterministic algorithm.

What carries the argument

Two-timescale asymptotic simulation of singularly perturbed hybrid inclusions, with limiting sets given by weakly invariant and internally chain-transitive sets of the boundary layer and reduced systems.

If this is right

Limiting behavior of the simulation is completely described by the weakly invariant and internally chain-transitive sets of the boundary layer and reduced systems.
A two-timescale stochastic approximation of a hybrid optimization algorithm recovers the trajectories of its deterministic counterpart under the given conditions.
The results apply directly to model-free analysis of convergence for any hybrid inclusion satisfying the structural assumptions.
Step-size sequences can be chosen so that the simulation property holds for a broad class of hybrid flow-jump systems.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same limiting-set characterization could be used to certify convergence rates once the boundary layer and reduced systems are explicitly solved.
The framework suggests a route to extend classical stochastic approximation results to hybrid systems whose jump maps are set-valued.
Numerical tests on concrete hybrid optimization problems could check whether observed stochastic trajectories remain close to the predicted deterministic limits for large iteration counts.

Load-bearing premise

The hybrid inclusion must admit well-defined boundary layer and reduced systems, and the sequences of iterates and step sizes must satisfy the listed sufficient conditions.

What would settle it

A concrete sequence of iterates and step sizes that meets the sufficient conditions yet whose omega-limit set lies outside the weakly invariant sets of the boundary layer and reduced systems, or a stochastic hybrid optimization example whose trajectories diverge from those of the deterministic version.

read the original abstract

Convergence properties of model-free two-timescale asymptotic simulations of singularly perturbed hybrid inclusions are developed. A hybrid inclusion combines constrained differential and difference inclusions to capture continuous (flow) and discrete (jump) dynamics, respectively. Sufficient conditions are established under which sequences of iterates and step sizes constitute a two-timescale asymptotic simulation of such a system, with limiting behavior characterized via weakly invariant and internally chain-transitive sets of an associated boundary layer and reduced system. To illustrate the applicability of these results, conditions are given under which a two-timescale stochastic approximation of a hybrid optimization algorithm asymptotically recovers the behavior of its deterministic counterpart.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper sets out sufficient conditions for two-timescale asymptotic simulations of singularly perturbed hybrid inclusions and shows how they let a stochastic approximation recover deterministic hybrid optimization behavior.

read the letter

The main point is that the authors supply sufficient conditions on iterates and step sizes so that they form a two-timescale asymptotic simulation of a hybrid inclusion, with the limit sets characterized by weakly invariant and internally chain-transitive sets of the boundary-layer and reduced systems. They then apply the same conditions to a stochastic approximation scheme for hybrid optimization and obtain asymptotic equivalence to the deterministic case. That combination of a general simulation result plus the optimization illustration is the concrete new piece. The argument follows the standard route of constructing the limiting systems from the hybrid structure and then using invariance principles, which keeps the logic transparent once the conditions are granted. The conditions themselves look workable on paper and avoid obvious circularity. The main limitation is that the conditions on the step-size sequences and the structural assumptions on the inclusion are fairly specific; it is not yet clear how often they hold for the kinds of noisy or model-free schemes people actually run. The optimization example is presented as an illustration rather than a worked numerical case, so the practical tightness of the result is still open. This is written for people already working in hybrid dynamical systems and stochastic approximation. A reader comfortable with Teel-style hybrid inclusions will follow the steps without trouble and can extract the two-timescale extension directly. The framework is grounded enough and the application direction relevant enough that it should go to peer review rather than a desk rejection.

Referee Report

0 major / 3 minor

Summary. The paper develops convergence properties of model-free two-timescale asymptotic simulations of singularly perturbed hybrid inclusions. A hybrid inclusion combines constrained differential and difference inclusions to capture continuous (flow) and discrete (jump) dynamics. Sufficient conditions are established under which sequences of iterates and step sizes constitute a two-timescale asymptotic simulation of such a system, with limiting behavior characterized via weakly invariant and internally chain-transitive sets of an associated boundary layer and reduced system. The results are illustrated by conditions under which a two-timescale stochastic approximation of a hybrid optimization algorithm asymptotically recovers the behavior of its deterministic counterpart.

Significance. If the sufficient conditions hold, the framework provides a rigorous, parameter-free approach to analyzing limits of two-timescale hybrid stochastic approximations via invariance principles on the boundary-layer and reduced systems. This extends existing asymptotic simulation techniques to hybrid inclusions and offers a concrete tool for verifying that stochastic hybrid optimization algorithms recover deterministic behavior, which is valuable for control and optimization applications involving mixed continuous-discrete dynamics.

minor comments (3)

The abstract and introduction would benefit from an explicit statement of the precise hybrid inclusion structure (e.g., the form of the flow and jump maps) used to construct the boundary-layer and reduced systems; this would help readers verify the structural assumptions without consulting external references.
Notation for the step-size sequences and the two time scales should be introduced with a single consolidated table or definition block early in the paper to avoid repeated redefinition across sections.
The application section on stochastic hybrid optimization would be strengthened by a brief remark on how the weakly invariant sets translate into practical convergence guarantees (e.g., to equilibria or cycles) for the optimization algorithm.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive assessment of the manuscript, accurate summary of the contributions on convergence properties for two-timescale asymptotic simulations of singularly perturbed hybrid inclusions, and recommendation of minor revision. The significance statement correctly highlights the value for control and optimization applications with mixed continuous-discrete dynamics.

Circularity Check

0 steps flagged

Minor self-citation present but not load-bearing

full rationale

The paper derives sufficient conditions for sequences of iterates and step sizes to form two-timescale asymptotic simulations of singularly perturbed hybrid inclusions, with limits characterized by weakly invariant and internally chain-transitive sets of boundary-layer and reduced systems. These conditions are constructed directly from the hybrid inclusion structure and standard invariance principles, without any reduction to fitted parameters, self-definitions, or ansatzes. Citations to prior hybrid systems results (including by the authors) supply background definitions but are not load-bearing for the central claims, which remain independently verifiable via the stated assumptions on flows, jumps, and step-size sequences. The derivation is self-contained against external mathematical benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Based on the abstract alone, no explicit free parameters, ad-hoc axioms, or invented entities are identified; the work relies on standard concepts from hybrid inclusions and singular perturbation theory.

pith-pipeline@v0.9.0 · 5407 in / 1079 out tokens · 36799 ms · 2026-05-13T18:54:28.022733+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

limiting behavior characterized via weakly invariant and internally chain-transitive sets of an associated boundary layer and reduced system

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

14 extracted references · 14 canonical work pages

[1]

Chain transitivity in generalized hybrid dynamics with application to simulation and stochastic approximation of hybrid systems,

R. K. Goebel and A. R. Teel, “Chain transitivity in generalized hybrid dynamics with application to simulation and stochastic approximation of hybrid systems,” 2026, arXiv Preprint

work page 2026
[2]

Asymptotic properties of asymptotic simulations of hybrid inclusions,

——, “Asymptotic properties of asymptotic simulations of hybrid inclusions,” 2026, submitted to IEEE CDC 2026

work page 2026
[3]

Asymptotic pseudotrajectories and chain recurrent flows, with applications,

M. Bena ¨ım and M. W. Hirsch, “Asymptotic pseudotrajectories and chain recurrent flows, with applications,”Journal of Dynamics and Differential Equations, vol. 8, no. 1, pp. 141–176, Jan. 1996

work page 1996
[4]

Stochastic Approximations and Differential Inclusions,

M. Bena¨ım, J. Hofbauer, and S. Sorin, “Stochastic Approximations and Differential Inclusions,”SIAM Journal on Control and Optimization, vol. 44, no. 1, pp. 328–348, Jan. 2005

work page 2005
[5]

Stochastic approxima- tion of hybrid systems: Boundedness and asymptotic behavior,

A. R. Teel, R. G. Sanfelice, and R. K. Goebel, “Stochastic approxima- tion of hybrid systems: Boundedness and asymptotic behavior,”Annual Reviews in Control, vol. 60, p. 101015, Jan. 2025

work page 2025
[6]

Stochastic approximation results for hybrid inclusions,

A. R. Teel, R. K. Goebel, R. G. Sanfelice, and M. F. Crisafulli, “Stochastic approximation results for hybrid inclusions,” in2024 IEEE 63rd Conference on Decision and Control (CDC), Dec. 2024, pp. 7822–7827

work page 2024
[7]

Stochastic approximation with two time scales,

V . S. Borkar, “Stochastic approximation with two time scales,”Systems & Control Letters, vol. 29, no. 5, pp. 291–294, Feb. 1997

work page 1997
[8]

Stochastic approximation with two time scales: The general case,

——, “Stochastic approximation with two time scales: The general case,”Stochastic Processes and their Applications, vol. 190, p. 104759, Dec. 2025

work page 2025
[9]

Hybrid Heavy-Ball Systems: Reset Methods for Optimization with Uncertainty,

J. H. Le and A. R. Teel, “Hybrid Heavy-Ball Systems: Reset Methods for Optimization with Uncertainty,” in2021 American Control Conference (ACC), May 2021, pp. 2236–2241

work page 2021
[10]

R. T. Rockafellar and R. J. B. Wets,Variational Analysis. Springer, 1998

work page 1998
[11]

Goebel, R

R. Goebel, R. G. Sanfelice, and A. R. Teel,Hybrid Dynamical Systems: Modeling, Stability, and Robustness. Princeton University Press, 2012

work page 2012
[12]

A direct proof of Conley’s decomposition for well-posed hybrid inclusions,

R. Goebel, “A direct proof of Conley’s decomposition for well-posed hybrid inclusions,”Systems & Control Letters, vol. 180, p. 105604, Oct. 2023

work page 2023
[13]

Systems of differential equations containing small parameters in the derivatives,

A. N. Tikhonov, “Systems of differential equations containing small parameters in the derivatives,”Matematicheskii Sbornik. Novaya Seriya, vol. 31(73), no. 3, pp. 575–586, 1952

work page 1952
[14]

A smooth Conley–Lyapunov function for hybrid inclusions on Rn,

R. K. Goebel and A. R. Teel, “A smooth Conley–Lyapunov function for hybrid inclusions on Rn,”Systems & Control Letters, vol. 204, p. 106204, Oct. 2025. APPENDIX Lemma A.1.Let the pair of step sizes {(hs,k, hf,k)}∞ k=1 be two-timescale admissible. Then, for each T >0 , there exists an ℓ∈Z ≥0 such that If,n,T and Is,n,T are nonempty and If,n,T ⊂ I s,n,T for...

work page 2025