On the suboptimality of stochastic MPC with varying constraint horizon

Allan Andre Do Nascimento; Andre Bertolace; Antonis Papachristodoulou; Kostas Margellos

Home/Archive/arXiv

arXiv · 2608.30017

On the suboptimality of stochastic MPC with varying constraint horizon

Allan Andre Do Nascimento·Andre Bertolace·Antonis Papachristodoulou·Kostas Margellos

View original source ↗Read source document ↗Export metadata ↓

Abstract

Enforcing stochastic state constraints over the full prediction horizon in Model Predictive Control (MPC) can be computationally demanding. Here we study stochastic MPC without terminal ingredients in which chance constraints are enforced only over a shorter constraint horizon. Using stochastic relaxed dynamic programming, we derive an explicit upper bound on the average expected closed-loop cost that depends on both prediction and constraint horizons. For linear quadratic problems with affine chance constraints and bounded uniform disturbances, we provide a deterministic reformulation via coordinate transformation and constraint tightening. Simulations illustrate the trade-off between computational effort and performance.

Explore related subjects

math.OC cs.SY eess.SY

Keep this discovery

BibTeX RIS

Allan Andre Do Nascimento, Andre Bertolace, Antonis Papachristodoulou, Kostas Margellos. 2026-08-30. On the suboptimality of stochastic MPC with varying constraint horizon. https://arxiv.org/abs/2608.30017

Cite the original work for its findings. Save a collection to share your selection of sources.

Discover connections

cs.SY subject eess.SY subject math.OC subject

Connections use source metadata and explicit phrase matches, not verified experimental comparisons.

KEEP EXPLORING

Related discoveries

arXiv2026

Rates of Convergence in the Central Limit Theorem for Markov Chains, with an Application to TD Learning

We prove a non-asymptotic central limit theorem for vector-valued martingale differences using Stein's method, and use Poisson's equation to extend the result to functions of Markov Chains. We then show that these results can be applied to establish a non-asymptotic central limit theorem for Temporal Difference (TD) learning with averaging.

math.PR↗

arXiv2026

Constrained Parameter Update Law for Adaptive Control

In this paper, constrained parameter update laws for adaptive control are developed using barrier constraints. An interpretation of the parameter update law from a constrained optimization problem, in which a regularized Barrier saddle function is formulated to incorporate parameter constraints using inverse and logarithmic barrier functions from interior-point methods. The resulting constrained update law is integrated with an adaptive trajectory tracking controller, enabling online learning of the unknown system model parameters. Forward invariance of the parameter estimate is established and Lyapunov stability of the closed-loop system with the constrained parameter update law is derived. The effectiveness of the proposed constrained adaptive control law is demonstrated through simulations, which validate its ability to maintain parameter estimates within prescribed bounds while ensuring convergence to the true parameter values and achieving steady state tracking performance.

math.OC↗

arXiv2026

Quiver Semistability and Structured Kalman Decompositions for Networked Linear Dynamical Systems

We introduce new notions of controllability and observability for networked linear time-invariant (LTI) systems based on $σ$-semistability of quiver representations. Utilizing King's criterion for $σ$-semistability, we define a network generalization of the Kalman decomposition for networked LTI systems, which systematically decomposes the local and interconnection dynamics while respecting the underlying network structure. Furthermore, we present efficient algorithms for deciding the proposed controllability and observability of a given networked LTI system and for finding the Kalman-type decomposition. We also show efficient algorithms for deciding the $σ$-semistability of representations of acyclic quivers with self-loops if the weight $σ$ has the same sign for all vertices with self-loops. Such quiver representations and weights arise from networked LTI systems.

math.OC↗