Scholarlink

<p>To enhance vehicle comfort and handling stability across diverse road conditions, this study proposes a novel adaptive extended Linear Quadratic Regulator (LQR) control strategy for semi-active suspension, integrating random forest-based road identification and multi-objective optimization. The extended LQR controller incorporates disturbance feedforward gain and employs a Multi-Objective Improved Genetic Algorithm (MIGA) to optimize weighting parameters. Hardware-in-the-loop (HIL) experiments validate the strategy’s performance. Under straight-line conditions, the proposed strategy reduces vehicle vertical acceleration by 35.5% and 13.6%, pitch angular acceleration by 14.9% and 13.6%, roll angular acceleration by 52.2% and 24.1%, suspension deflection by 10.2% and 14.9%, and tire dynamic load by 3.0% compared to passive suspension and traditional LQR control, respectively. Although tire dynamic load is slightly higher than passive suspension, it achieves a balanced optimization of comprehensive performance. Under lane-change conditions, improvements include 25.5% and 12.6% in vertical acceleration, 24.1% and 14.1% in pitch angular acceleration, and 45.5% and 25.8% in roll angular acceleration, respectively. These results confirm the strategy’s robust dynamic adaptability and superior control performance.</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

Study on Adaptive Extended LQR Control for Suspension Based on Road Recognition with Experimental Validation

  • 期刊论文

  • 发表日期：

    2026年3月30日

查看全文
  • Qingyang Zhao,
  • Yaohui Lu,
  • Lijiang Zhu,
  • Shiying Li,
  • Haoyuan Wang,
  • Yuanyuan Ma,
  • Jun Xu

摘要

To enhance vehicle comfort and handling stability across diverse road conditions, this study proposes a novel adaptive extended Linear Quadratic Regulator (LQR) control strategy for semi-active suspension, integrating random forest-based road identification and multi-objective optimization. The extended LQR controller incorporates disturbance feedforward gain and employs a Multi-Objective Improved Genetic Algorithm (MIGA) to optimize weighting parameters. Hardware-in-the-loop (HIL) experiments validate the strategy’s performance. Under straight-line conditions, the proposed strategy reduces vehicle vertical acceleration by 35.5% and 13.6%, pitch angular acceleration by 14.9% and 13.6%, roll angular acceleration by 52.2% and 24.1%, suspension deflection by 10.2% and 14.9%, and tire dynamic load by 3.0% compared to passive suspension and traditional LQR control, respectively. Although tire dynamic load is slightly higher than passive suspension, it achieves a balanced optimization of comprehensive performance. Under lane-change conditions, improvements include 25.5% and 12.6% in vertical acceleration, 24.1% and 14.1% in pitch angular acceleration, and 45.5% and 25.8% in roll angular acceleration, respectively. These results confirm the strategy’s robust dynamic adaptability and superior control performance.