Scholarlink

This article focuses on the selection of reactive current ratio coefficient, explaining the impact of fixed current coefficient on wind turbine voltage recovery from the perspectives of maximizing dynamic reactive power capacity during faults and extreme voltage surges during fault recovery, which bring high voltage ride through risks. Subsequently, the establishment of a dynamic current coefficient adjustment model was completed within the framework of reinforcement learning, and reasonable modifications were made to the algorithm mechanism to address the problem. The PPO algorithm was selected for intelligent agent training in both continuous state space and discrete action space. Finally, the superiority of the algorithm and the effectiveness of improving wind turbine transient voltage support and accelerating transient voltage recovery by changing the reactive current ratio coefficient were verified in simulation testing. The variable reactive current ratio coefficient can achieve maximum reactive power support during the duration of faults, and better tap into the reactive capacity of doubly fed wind turbines.

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

Reactive Current Proportional Coefficient Adjustment Based on Reinforcement Learning

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Qing Mengqi,
  • Huang Da,
  • Wang Xiyue,
  • Hu Guoli,
  • Liu Yongyan,
  • Ke Deping

摘要

This article focuses on the selection of reactive current ratio coefficient, explaining the impact of fixed current coefficient on wind turbine voltage recovery from the perspectives of maximizing dynamic reactive power capacity during faults and extreme voltage surges during fault recovery, which bring high voltage ride through risks. Subsequently, the establishment of a dynamic current coefficient adjustment model was completed within the framework of reinforcement learning, and reasonable modifications were made to the algorithm mechanism to address the problem. The PPO algorithm was selected for intelligent agent training in both continuous state space and discrete action space. Finally, the superiority of the algorithm and the effectiveness of improving wind turbine transient voltage support and accelerating transient voltage recovery by changing the reactive current ratio coefficient were verified in simulation testing. The variable reactive current ratio coefficient can achieve maximum reactive power support during the duration of faults, and better tap into the reactive capacity of doubly fed wind turbines.