Scholarlink

This paper presents a deep reinforcement learning (DRL)-based synchronization method for the Hindmarsh-Rose (H-R) model systems. The H-R model is a well-established nonlinear system that accurately captures the dynamics of neuronal bursting. The paper describes how the proximal policy optimization (PPO) algorithm, a variant of the trust region policy optimization (TRPO) method, is utilized to achieve synchronization between two H-R models. The approach requires only two control inputs and demonstrates robust performance even under random initial conditions and in the presence of noise. Simulation results show that the proposed method can effectively synchronize the H-R models with low error and good noise rejection capability. The DRL-based method does not require a priori knowledge of the system dynamics, making it a versatile solution for chaos control problems.

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

Synchronization Between Two Hindmarsh-Rose Neural Models via Deep Reinforcement Learning Method

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Shitao Jin,
  • Jie Wu

摘要

This paper presents a deep reinforcement learning (DRL)-based synchronization method for the Hindmarsh-Rose (H-R) model systems. The H-R model is a well-established nonlinear system that accurately captures the dynamics of neuronal bursting. The paper describes how the proximal policy optimization (PPO) algorithm, a variant of the trust region policy optimization (TRPO) method, is utilized to achieve synchronization between two H-R models. The approach requires only two control inputs and demonstrates robust performance even under random initial conditions and in the presence of noise. Simulation results show that the proposed method can effectively synchronize the H-R models with low error and good noise rejection capability. The DRL-based method does not require a priori knowledge of the system dynamics, making it a versatile solution for chaos control problems.