Scholarlink

<p>This paper proposes a novel decoupling algorithm with an error auto-compensation strategy for general six-axis acceleration sensing mechanisms. First, mathematical preliminaries on vectors, matrices and quaternions are given. Then, dynamic equations are derived according to Kane’s equations and solved numerically using the trapezoid formula. To improve the accuracy of the solution, an auto-compensation strategy based on the vibration properties is introduced. Additionally, the compensation conditions of this strategy are determined through error characteristic analysis. Subsequently, virtual prototype experiments are conducted to verify the performances of the proposed algorithm. The results indicate that: (1) under 5% interference noise, the proposed algorithm performs effectively across all chosen configurations (“6–6”, “9–3”, “9–4”, “12–4” and “12–6”), with comprehensive relative errors reduced by up to 54.15%; (2) it takes the proposed algorithm only 0.13&#xa0;s to process data spanning 3&#xa0;s, with a comprehensive relative error of 0.03%; (3) despite the varying interference noise, the comprehensive relative error remains within 1.06%. Finally, the actual prototype experiment further demonstrates the feasibility of the proposed decoupling algorithm, with a maximum comprehensive relative error of 5.82%.</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

A novel decoupling algorithm covering error auto-compensation for general six-axis acceleration sensing mechanisms

  • 期刊论文

  • 发表日期：

    2026年1月12日

查看全文
  • Yuanwei Zhang,
  • Jingjing You,
  • Pengda Ye,
  • Xianzhu Zhang,
  • Jie Hua

摘要

This paper proposes a novel decoupling algorithm with an error auto-compensation strategy for general six-axis acceleration sensing mechanisms. First, mathematical preliminaries on vectors, matrices and quaternions are given. Then, dynamic equations are derived according to Kane’s equations and solved numerically using the trapezoid formula. To improve the accuracy of the solution, an auto-compensation strategy based on the vibration properties is introduced. Additionally, the compensation conditions of this strategy are determined through error characteristic analysis. Subsequently, virtual prototype experiments are conducted to verify the performances of the proposed algorithm. The results indicate that: (1) under 5% interference noise, the proposed algorithm performs effectively across all chosen configurations (“6–6”, “9–3”, “9–4”, “12–4” and “12–6”), with comprehensive relative errors reduced by up to 54.15%; (2) it takes the proposed algorithm only 0.13 s to process data spanning 3 s, with a comprehensive relative error of 0.03%; (3) despite the varying interference noise, the comprehensive relative error remains within 1.06%. Finally, the actual prototype experiment further demonstrates the feasibility of the proposed decoupling algorithm, with a maximum comprehensive relative error of 5.82%.