Safe Learning-Based Control of Stochastic Jump Linear Systems: a Distributionally Robust Approach

被引:0
作者
Schuurmans, Mathijs [1 ]
Sopasakis, Pantelis [2 ]
Patrinos, Panagiotis [1 ]
机构
[1] Katholieke Univ Leuven, Dept Elect Engn ESAT TADIUS, Kasteelpk Arenberg 10, B-3001 Leuven, Belgium
[2] Queens Univ Belfast, Sch Elect Elect Engn & Comp Sci, Ctr Intelligent Autonomous Mfg Syst, Belfast BT9 5AH, Antrim, North Ireland
来源
2019 IEEE 58TH CONFERENCE ON DECISION AND CONTROL (CDC) | 2019年
关键词
MODEL PREDICTIVE CONTROL; OPTIMIZATION;
D O I
10.1109/cdc40024.2019.9029946
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
We consider the problem of designing control laws for stochastic jump linear systems where the disturbances are drawn randomly from a finite sample space according to an unknown distribution, which is estimated from a finite sample of i.i.d. observations. We adopt a distributionally robust approach to compute a mean-square stabilizing feedback gain with a given probability. The larger the sample size, the less conservative the controller, yet our methodology gives stability guarantees with high probability, for any number of samples. Using tools from statistical learning theory, we estimate confidence regions for the unknown probability distributions (ambiguity sets) which have the shape of total variation balls centered around the empirical distribution. We use these confidence regions in the design of appropriate distributionally robust controllers and show that the associated stability conditions can be cast as a tractable linear matrix inequality (LMI) by using conjugate duality. The resulting design procedure scales gracefully with the size of the probability space and the system dimensions. Through a numerical example, we illustrate the superior sample complexity of the proposed methodology over the stochastic approach.
引用
收藏
页码:6498 / 6503
页数:6
相关论文
共 28 条
[1]  
[Anonymous], 2017, MOSEK OPTIMIZATION T
[2]  
Bauschke HH, 2011, CMS BOOKS MATH, P1, DOI 10.1007/978-1-4419-9467-7
[3]   Robust Solutions of Optimization Problems Affected by Uncertain Probabilities [J].
Ben-Tal, Aharon ;
den Hertog, Dick ;
De Waegenaere, Anja ;
Melenberg, Bertrand ;
Rennen, Gijs .
MANAGEMENT SCIENCE, 2013, 59 (02) :341-357
[4]   Data-driven robust optimization [J].
Bertsimas, Dimitris ;
Gupta, Vishal ;
Kallus, Nathan .
MATHEMATICAL PROGRAMMING, 2018, 167 (02) :235-292
[5]  
Boucheron S., 2013, Concentration inequalities: A nonasymptotic theory of independence, DOI DOI 10.1093/ACPROF:OSO/9780199535255.001.0001
[6]  
Boyd S., 1994, SIAM studies in applied mathematics
[7]  
Costa O. L. V., 2006, Discrete-Time Markov Jump Linear Systems
[8]   Distributionally Robust Optimization Under Moment Uncertainty with Application to Data-Driven Problems [J].
Delage, Erick ;
Ye, Yinyu .
OPERATIONS RESEARCH, 2010, 58 (03) :595-612
[9]  
Domahidi A, 2011, IEEE DECIS CONTR P, P513, DOI 10.1109/CDC.2011.6161258
[10]  
Dupacova J., 1987, Stochastics, V20, P73, DOI 10.1080/17442508708833436