Optimal Control of Two-Dimensional Roesser Model: Solution Based on Reinforcement Learning

• 成果类型: Article
• 署名作者: Ye, Linwei; Zhao, Zhonggai; Liu, Fei
• 署名单位: Jiangnan University; Jiangnan University
• 刊物名称: IEEE TRANSACTIONS ON AUTOMATIC CONTROL
• ISSN/ISSBN: 0018-9286
• DOI: 10.1109/TAC.2024.3357743
• 发表日期: 2024
• 页码: 5424-5430
• 关键词: mathematical models; State feedback; optimal control; Couplings; Two-dimensional displays; trajectory; Reinforcement Learning; Discrete two-dimensional (2-D) Roesser model; infinite-region linear quadratic regulation (LQR); policy-iteration; value-iteration

摘要

This article addresses the infinite-region linear quadratic regulation problem of the discrete two-dimensional (2-D) Roesser model, drawing inspiration from reinforcement learning principles. It introduces a novel proof establishing that expressing the optimal control law in 2-D through state feedback is unattainable. Subsequently, a policy iteration framework is proposed to derive suboptimal state feedback, followed by an exploration of the true optimal policy through value iteration. The efficacy of these methodologies is demonstrated via two numerical examples, enhancing the practical understanding of these innovative approaches.