Learning-Based Distributionally Robust Model Predictive Control of\n Markovian Switching Systems with Guaranteed Stability and Recursive\n Feasibility

2020/09/09 by Mathijs Schuurmans, Panagiotis Patrinos, Schuurmans, Mathijs +1 · 1 citation
Engineering · #Advanced Control Systems Optimization #FOS: Electrical engineering #FOS: Mathematics #Optimization and Control (math.OC) #Systems and Control (eess.SY) #electronic engineering #information engineering

paper · pdf · doi:10.48550/arxiv.2009.04422

openalex publication_date 2020/09/09 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

We present a data-driven model predictive control scheme for\nchance-constrained Markovian switching systems with unknown switching\nprobabilities. Using samples of the underlying Markov chain, ambiguity sets of\ntransition probabilities are estimated which include the true conditional\nprobability distributions with high probability. These sets are updated online\nand used to formulate a time-varying, risk-averse optimal control problem. We\nprove recursive feasibility of the resulting MPC scheme and show that the\noriginal chance constraints remain satisfied at every time step. Furthermore,\nwe show that under sufficient decrease of the confidence levels, the resulting\nMPC scheme renders the closed-loop system mean-square stable with respect to\nthe true-but-unknown distributions, while remaining less conservative than a\nfully robust approach.\n

Cited by

• Bayesian Risk-averse Model Predictive Control with Consistency and Stability Guarantees

Related

• A General Framework for Learning-Based Distributionally Robust MPC of Markov Jump Systems
• Bayesian Risk-averse Model Predictive Control with Consistency and Stability Guarantees
• Stochastic economic model predictive control for Markovian switching systems
• Conformal Prediction-Based MPC for Stochastic Linear Systems
• Data-driven Distributionally Robust MPC: An indirect feedback approach