Scholarlink

<p>The current collaborative robots are no longer able to meet people’s demand for high-precision trajectory tracking. Therefore, this study proposes a dynamic motion kernel primitive algorithm that combines the kernel motion primitive algorithm and dynamic motion primitive algorithm. An adaptive neural network is introduced to design a trajectory-tracking controller. Algorithm validation showed that the research algorithm outperformed traditional algorithms in terms of running time, tracking trajectory error, and average reward return. On average, the convergence steps for three-letter trajectory tracking were reduced by 11.54% compared to other algorithms. The controller simulation experiment showed that the proposed controller reduced the trajectory tracking error of the end effector by 65.43% compared to the position controller and 53.33% compared to the adaptive controller. The results show that the proposed dynamic motion kernel primitive algorithm can effectively modulate the overall trajectory and improve trajectory tracking accuracy. The proposed controller has superiority in trajectory tracking and control of high stiffness flexible joint collaborative robots, and can effectively achieve high-precision tracking of collaborative robots.</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

Collaborative robot trajectory tracking control based on DS-KMP algorithm

  • 期刊论文

  • 发表日期：

    2025年2月28日

查看全文
  • Yan Liu,
  • Yixin Cao,
  • Chunmin Jiang

摘要

The current collaborative robots are no longer able to meet people’s demand for high-precision trajectory tracking. Therefore, this study proposes a dynamic motion kernel primitive algorithm that combines the kernel motion primitive algorithm and dynamic motion primitive algorithm. An adaptive neural network is introduced to design a trajectory-tracking controller. Algorithm validation showed that the research algorithm outperformed traditional algorithms in terms of running time, tracking trajectory error, and average reward return. On average, the convergence steps for three-letter trajectory tracking were reduced by 11.54% compared to other algorithms. The controller simulation experiment showed that the proposed controller reduced the trajectory tracking error of the end effector by 65.43% compared to the position controller and 53.33% compared to the adaptive controller. The results show that the proposed dynamic motion kernel primitive algorithm can effectively modulate the overall trajectory and improve trajectory tracking accuracy. The proposed controller has superiority in trajectory tracking and control of high stiffness flexible joint collaborative robots, and can effectively achieve high-precision tracking of collaborative robots.