Scholarlink

<p>The design of robust control systems for rehabilitation exoskeletons is essential to improve patient mobility and recovery. This paper proposes a novel model-free adaptive backstepping control strategy, specifically developed for a 10-degree-of-freedom (DOF) lower limb exoskeleton intended for rehabilitation. The originality of this work lies in the synergistic combination of several advanced techniques: a second-order ultra-local model is used to bypass the complexities of system dynamics, a multilayer perceptron (MLP) neural network is integrated to estimate and compensate for lumped disturbances in real time, and particle swarm optimization (PSO) is applied to automatically tune the controller parameters for optimal performance. This integrated approach enables adaptive, precise, and stable control without relying on a predefined model. The closed-loop system stability is formally proven using Lyapunov theory. Validation is carried out through co-simulation with SolidWorks, Simscape Multibody, and MATLAB/Robotics Toolbox, and the results demonstrate clear improvements in control accuracy and robustness compared to existing model-free methods.</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

Adaptive model-free control of lower limb exoskeletons using neural estimation and swarm-based optimization

  • 期刊论文

  • 发表日期：

    2025年10月3日

查看全文
  • Yasmine Amara,
  • Farid Kenas

摘要

The design of robust control systems for rehabilitation exoskeletons is essential to improve patient mobility and recovery. This paper proposes a novel model-free adaptive backstepping control strategy, specifically developed for a 10-degree-of-freedom (DOF) lower limb exoskeleton intended for rehabilitation. The originality of this work lies in the synergistic combination of several advanced techniques: a second-order ultra-local model is used to bypass the complexities of system dynamics, a multilayer perceptron (MLP) neural network is integrated to estimate and compensate for lumped disturbances in real time, and particle swarm optimization (PSO) is applied to automatically tune the controller parameters for optimal performance. This integrated approach enables adaptive, precise, and stable control without relying on a predefined model. The closed-loop system stability is formally proven using Lyapunov theory. Validation is carried out through co-simulation with SolidWorks, Simscape Multibody, and MATLAB/Robotics Toolbox, and the results demonstrate clear improvements in control accuracy and robustness compared to existing model-free methods.