Scholarlink

<p>This paper proposes a novel data-driven control method for a robotic mechanism inspired by caterpillar locomotion. A caterpillar robot is designed by accurately modeling its complex and nonlinear motion dynamics. To handle these nonlinearities, Koopman operator theory is employed to derive a linear data-driven dynamic model. A fractional-order sliding-mode controller is then developed to regulate the linearized system effectively. Considering the frequent external disturbances encountered by the caterpillar robot, an adaptive Radial Basis Function neural network is integrated to estimate these disturbances and mitigate the chattering phenomenon. Simulation results demonstrate that the proposed control method achieves high tracking accuracy with reduced chattering, confirming its effectiveness and potential for practical applications in robotic systems requiring robust and precise control. This work underscores the advantages of integrating Koopman theory, fractional-order sliding-mode control, and adaptive neural networks for bio-inspired robotic mechanisms.</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

Data-driven deep neural control for chattering reduction in a novel caterpillar robot mechanism

  • 期刊论文

  • 发表日期：

    2025年7月4日

查看全文
  • Mehran Rahmani,
  • Apoorva Rahul Uplap,
  • Sangram Redkar

摘要

This paper proposes a novel data-driven control method for a robotic mechanism inspired by caterpillar locomotion. A caterpillar robot is designed by accurately modeling its complex and nonlinear motion dynamics. To handle these nonlinearities, Koopman operator theory is employed to derive a linear data-driven dynamic model. A fractional-order sliding-mode controller is then developed to regulate the linearized system effectively. Considering the frequent external disturbances encountered by the caterpillar robot, an adaptive Radial Basis Function neural network is integrated to estimate these disturbances and mitigate the chattering phenomenon. Simulation results demonstrate that the proposed control method achieves high tracking accuracy with reduced chattering, confirming its effectiveness and potential for practical applications in robotic systems requiring robust and precise control. This work underscores the advantages of integrating Koopman theory, fractional-order sliding-mode control, and adaptive neural networks for bio-inspired robotic mechanisms.