Model-Free Design of Stochastic LQR Controller from Reinforcement Learning and Primal-Dual Optimization Perspective

• Published in: arXiv.org
• Main Authors: Li, Man, Qin, Jiahu, Wei Xing Zheng, Wang, Yaonan, Kang, Yu
• Format: Paper
• Language: English
• Published: Ithaca Cornell University Library, arXiv.org 17.03.2021
• Subjects: Algorithms; Computational geometry; Control systems design; Controllers; Convex analysis; Convexity; Iterative methods; Kuhn-Tucker method; Linear quadratic regulator; Linear systems; Machine learning; Optimization
• ISSN: 2331-8422

To further understand the underlying mechanism of various reinforcement learning (RL) algorithms and also to better use the optimization theory to make further progress in RL, many researchers begin to revisit the linear-quadratic regulator (LQR) problem, whose setting is simple and yet captures the characteristics of RL. Inspired by this, this work is concerned with the model-free design of stochastic LQR controller for linear systems subject to Gaussian noises, from the perspective of both RL and primal-dual optimization. From the RL perspective, we first develop a new model-free off-policy policy iteration (MF-OPPI) algorithm, in which the sampled data is repeatedly used for updating the policy to alleviate the data-hungry problem to some extent. We then provide a rigorous analysis for algorithm convergence by showing that the involved iterations are equivalent to the iterations in the classical policy iteration (PI) algorithm. From the perspective of optimization, we first reformulate the stochastic LQR problem at hand as a constrained non-convex optimization problem, which is shown to have strong duality. Then, to solve this non-convex optimization problem, we propose a model-based primal-dual (MB-PD) algorithm based on the properties of the resulting Karush-Kuhn-Tucker (KKT) conditions. We also give a model-free implementation for the MB-PD algorithm by solving a transformed dual feasibility condition. More importantly, we show that the dual and primal update steps in the MB-PD algorithm can be interpreted as the policy evaluation and policy improvement steps in the PI algorithm, respectively. Finally, we provide one simulation example to show the performance of the proposed algorithms.