Scholarlink

<p>This study presents a novel Forward Dynamic approach that integrates a Genetic Algorithm (GA) to generate and optimize gait patterns with Inverse Dynamics analysis, tailored to individual users in the robot-assisted training platform. The resulting trajectories provide gait-trajectory planning using the developed gait-pattern simulator, enabling medical personnel to select customized tasks and training trajectories for users based on their training goals, subsequently transferring user-specific parameters to the exoskeletal robot for rehabilitation/training. By leveraging biomechanical data as reference (joint torque and force) from the Inverse Dynamic analysis, the method foregoes the necessity for experimental data, directly predicting joint angles and positions. To ensure the validity of the proposed method, we used a combined approach of numerical analysis and comparison with motion-capture data. This evaluation aimed to assess how closely the simulated results resembled real human walking motion for the lower-extremity joints.</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

Tailored gait-pattern generation and verification based on the dynamic analysis of the human musculoskeletal model

  • 期刊论文

  • 发表日期：

    2025年1月8日

查看全文
  • Shamanth Shanmuga Prasad,
  • Youngwoo Kim

摘要

This study presents a novel Forward Dynamic approach that integrates a Genetic Algorithm (GA) to generate and optimize gait patterns with Inverse Dynamics analysis, tailored to individual users in the robot-assisted training platform. The resulting trajectories provide gait-trajectory planning using the developed gait-pattern simulator, enabling medical personnel to select customized tasks and training trajectories for users based on their training goals, subsequently transferring user-specific parameters to the exoskeletal robot for rehabilitation/training. By leveraging biomechanical data as reference (joint torque and force) from the Inverse Dynamic analysis, the method foregoes the necessity for experimental data, directly predicting joint angles and positions. To ensure the validity of the proposed method, we used a combined approach of numerical analysis and comparison with motion-capture data. This evaluation aimed to assess how closely the simulated results resembled real human walking motion for the lower-extremity joints.