Scholarlink

<p>This paper presents an intelligent adaptive formation control strategy for uncertain tilting quadrotors by integrating an Output-Recurrent Fuzzy Broad Learning System (ORFBLS), Reinforcement Learning (RL), and Adaptive Backstepping Sliding Mode Control (ABSMFC). The proposed hybrid method, abbreviated as ORFBLS-RL-ABSMFC, enables real-time trajectory tracking and robust consensus maintenance in dynamic multi-agent UAV environments. ORFBLS is employed to approximate system uncertainties with output feedback, while RL optimizes learning performance online. ABSMFC ensures system stability and convergence via Lyapunov-based design. The proposed ORFBLS-RL-ABSMF addresses practical challenges including actuator faults, mass variation, and external disturbances. Simulation and experimental results demonstrate superior adaptability, convergence speed, and disturbance rejection compared to existing control methods. Beyond UAV formation control, the raised framework offers potential applications to broader intelligent multi-agent formation problems in robotics, transportation, and autonomous systems.</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

Intelligent Adaptive Formation Control Using ORFBLS and Reinforcement Learning for Uncertain Tilting Multi-Quadrotors

  • 期刊论文

  • 发表日期：

    2025年10月3日

查看全文
  • Ching-Chih Tsai,
  • Chun-Fu Mao,
  • Kumail Hussain

摘要

This paper presents an intelligent adaptive formation control strategy for uncertain tilting quadrotors by integrating an Output-Recurrent Fuzzy Broad Learning System (ORFBLS), Reinforcement Learning (RL), and Adaptive Backstepping Sliding Mode Control (ABSMFC). The proposed hybrid method, abbreviated as ORFBLS-RL-ABSMFC, enables real-time trajectory tracking and robust consensus maintenance in dynamic multi-agent UAV environments. ORFBLS is employed to approximate system uncertainties with output feedback, while RL optimizes learning performance online. ABSMFC ensures system stability and convergence via Lyapunov-based design. The proposed ORFBLS-RL-ABSMF addresses practical challenges including actuator faults, mass variation, and external disturbances. Simulation and experimental results demonstrate superior adaptability, convergence speed, and disturbance rejection compared to existing control methods. Beyond UAV formation control, the raised framework offers potential applications to broader intelligent multi-agent formation problems in robotics, transportation, and autonomous systems.