Reinforcement learning-based composite suboptimal control for Markov
jump singularly perturbed systems with unknown dynamics
- Jiacheng Wu,
- Wenqian Li,
- Yun Wang,
- Hao Shen
Jiacheng Wu
Anhui University of Technology
Corresponding Author:jc980507@163.com
Author ProfileAbstract
In this article, a model-free parallel reinforcement learning method is
proposed to solve the suboptimal control problem for the Markov jump
singularly perturbed systems. First, since fast and slow dynamics
coexist in Markov jump singularly perturbed systems, it may lead to
ill-conditioned numerical problems during the controller design process.
Therefore, the original system can be decomposed into independent
subsystems at different time-scales by employing the reduced order
method. Besides, a model-based parallel algorithm is designed to obtain
the optimal controllers of the fast and slow subsystems respectively.
Moreover, within the framework of reinforcement learning, the composite
controller of the Markov jump singularly perturbed systems can be
obtained without system dynamics. Finally, a numerical example is
introduced to prove the effectiveness of proposed algorithms.28 Jul 2023Submitted to Optimal Control, Applications and Methods 28 Jul 2023Submission Checks Completed
28 Jul 2023Assigned to Editor
28 Jul 2023Review(s) Completed, Editorial Evaluation Pending
18 Aug 2023Reviewer(s) Assigned