Scholarlink

This article introduces a model predictive control method for closed-chain robots. By utilizing virtual breakpoints, this method simplifies complex closed-chain structures into more easily controllable open chains, facilitating standard kinematic analysis and the application of Model Predictive Control (MPC). The method incorporates the Iterative Linear Quadratic Regulator (iLQR), an effective Hessian approximation variant of Differential Dynamic Programming (DDP), known for its ability to handle discrete-time optimal control challenges efficiently. This technique integrates geometric constraints with dynamic simulation through the OCS2 framework to implement ILQR-MPC, ensuring precise motion prediction and enhancing real-time performance, which is crucial for dynamic systems. Initial theoretical and simulation validations on the Gazebo platform show promising results, with plans for further research to refine and validate this methodology. This innovative approach provides a robust and adaptable solution for controlling closed-chain mechanisms, significantly improving their design and operational capabilities.

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

Advancing Closed-Chain Robot Control Through Model Predictive Techniques and Virtual Breakpoints

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Jiayang Zhang,
  • Huashan Feng,
  • Yebiao Zhang,
  • Lefan Kang

摘要

This article introduces a model predictive control method for closed-chain robots. By utilizing virtual breakpoints, this method simplifies complex closed-chain structures into more easily controllable open chains, facilitating standard kinematic analysis and the application of Model Predictive Control (MPC). The method incorporates the Iterative Linear Quadratic Regulator (iLQR), an effective Hessian approximation variant of Differential Dynamic Programming (DDP), known for its ability to handle discrete-time optimal control challenges efficiently. This technique integrates geometric constraints with dynamic simulation through the OCS2 framework to implement ILQR-MPC, ensuring precise motion prediction and enhancing real-time performance, which is crucial for dynamic systems. Initial theoretical and simulation validations on the Gazebo platform show promising results, with plans for further research to refine and validate this methodology. This innovative approach provides a robust and adaptable solution for controlling closed-chain mechanisms, significantly improving their design and operational capabilities.