Scholarlink

<p>Ferromagnetic continuum robots (FCRs), renowned for their high flexibility and infinite degrees of freedom, represent a cutting-edge tool in medical applications. Despite their potential, designing a model-based controller with real-time solvability and high precision poses significant challenges in complex medical environments due to their nonlinear and magnetic properties. This paper pioneers a robust model-based controller design for FCRs by integrating kinematic and dynamic modeling to achieve high precision with real-time solvability, essential for high-precision medical tasks. The proposed controller is designed in two stages: first, a neural network maps positions to the desired strain based on the robot's kinematic structure, using training data derived from the kinematic model. Second, a nonlinear sliding mode controller ensures high stability and accuracy in regulating the tip position while precisely tracking specified points and trajectories. Comprehensive performance evaluations through simulations including point-to-point position control and trajectory tracking demonstrate the controller's efficacy. A comparison with experimental data for two robots reveals that the proposed approach achieves target positions with high precision, maintaining errors below 2% of the robots' length. Overall, the proposed controller demonstrates a strong alignment between theoretical and practical results, confirming its robustness, high precision, and stability for critical medical applications, making it a reliable solution for high-precision tasks in complex environments.</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

Dynamic model-based design of a hybrid neural network and sliding mode controller for real-time and accurate control of ferromagnetic continuum robots

  • 期刊论文

  • 发表日期：

    2025年8月29日

查看全文
  • Pouya Mallahi Kolahi,
  • Moharram Habibnejad Korayem

摘要

Ferromagnetic continuum robots (FCRs), renowned for their high flexibility and infinite degrees of freedom, represent a cutting-edge tool in medical applications. Despite their potential, designing a model-based controller with real-time solvability and high precision poses significant challenges in complex medical environments due to their nonlinear and magnetic properties. This paper pioneers a robust model-based controller design for FCRs by integrating kinematic and dynamic modeling to achieve high precision with real-time solvability, essential for high-precision medical tasks. The proposed controller is designed in two stages: first, a neural network maps positions to the desired strain based on the robot's kinematic structure, using training data derived from the kinematic model. Second, a nonlinear sliding mode controller ensures high stability and accuracy in regulating the tip position while precisely tracking specified points and trajectories. Comprehensive performance evaluations through simulations including point-to-point position control and trajectory tracking demonstrate the controller's efficacy. A comparison with experimental data for two robots reveals that the proposed approach achieves target positions with high precision, maintaining errors below 2% of the robots' length. Overall, the proposed controller demonstrates a strong alignment between theoretical and practical results, confirming its robustness, high precision, and stability for critical medical applications, making it a reliable solution for high-precision tasks in complex environments.