Scholarlink

<p>In this paper, a data-driven solution for the linear quadratic tracking (LQT) problem in affine linear parameter-varying (LPV) systems is presented. By solving a data-driven convex optimization problem, a common Lyapunov function is presented for the entire region of the system parameter changes. Two off-policy and on-policy algorithms are presented to solve the problem. Proof of stability and convergence of the algorithm is provided. Finally, The performance of the given algorithms for controlling an air vehicle and a single-link flexible joint robot has been simulated and evaluated.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Linear Quadratic Tracking Control of Continuous-time Affine Linear Parameter-varying Systems Using Reinforcement Learning

  • 期刊论文

  • 发表日期：

    2025年8月13日

查看全文
  • Hosein Ranjbarpur,
  • Mohammad Bagher Menhaj,
  • Hajar Atrianfar

摘要

In this paper, a data-driven solution for the linear quadratic tracking (LQT) problem in affine linear parameter-varying (LPV) systems is presented. By solving a data-driven convex optimization problem, a common Lyapunov function is presented for the entire region of the system parameter changes. Two off-policy and on-policy algorithms are presented to solve the problem. Proof of stability and convergence of the algorithm is provided. Finally, The performance of the given algorithms for controlling an air vehicle and a single-link flexible joint robot has been simulated and evaluated.