2021/05/06 by Mathijs Schuurmans, Panagiotis Patrinos, Schuurmans, Mathijs +1 · 1 citation
Engineering · #Advanced Control Systems Optimization #Control Systems and Identification #FOS: Electrical engineering #FOS: Mathematics #Fault Detection and Control Systems #Optimization and Control (math.OC) #Systems and Control (eess.SY) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2105.02511
openalex publication_date 2021/05/06 · openalex created_date 2022/07/25 · openalex updated_date 2026/07/28
We study safe, data-driven control of (Markov) jump linear systems with\nunknown transition probabilities, where both the discrete mode and the\ncontinuous state are to be inferred from output measurements. To this end, we\ndevelop a receding horizon estimator which uniquely identifies a sub-sequence\nof past mode transitions and the corresponding continuous state, allowing for\narbitrary switching behavior. Unlike traditional approaches to mode estimation,\nwe do not require an offline exhaustive search over mode sequences to determine\nthe size of the observation window, but rather select it online. If the system\nis weakly mode observable, the window size will be upper bounded, leading to a\nfinite-memory observer. We integrate the estimation procedure with a simple\ndistributionally robust controller, which hedges against misestimations of the\ntransition probabilities due to finite sample sizes. As additional mode\ntransitions are observed, the used ambiguity sets are updated, resulting in\ncontinual improvements of the control performance. The practical applicability\nof the approach is illustrated on small numerical examples.\n