欢迎访问《汽车安全与节能学报》，
汽车安全与节能学报

汽车安全与节能学报 ›› 2025, Vol. 16 ›› Issue (4): 598-609.DOI: 10.3969/j.issn.1674-8484.2025.04.010

• 智能驾驶与智慧交通 • 上一篇    下一篇

基于模糊MPC的无人矿卡横向控制

宗健壮(), 吴光强(), 毛礼波, 桂雨晖   

  1. 同济大学汽车学院， 上海 201804， 中国
  • 收稿日期:2025-01-19 修回日期:2025-03-03 出版日期:2025-08-30 发布日期:2025-08-27
  • 通讯作者: *吴光强，教授。E-mail：wuguangqiang@tongji.edu.cn。
  • 作者简介:宗健壮（1999—），男（汉），安徽，硕士研究生。E-mail：2233557@tongji.edu.cn。
  • 基金资助:
    特种车辆智能驾驶联合实验室横纵向控制算法开发项目(17002370097)

Lateral control for unmanned mining trucks based on fuzzy MPC

ZONG Jianzhuang(), WU Guangqiang(), MAO Libo, GUI Yuhui   

  1. School of Automotive Studies, Tongji University, Shanghai 201800, China
  • Received:2025-01-19 Revised:2025-03-03 Online:2025-08-30 Published:2025-08-27

RichHTML

4

PDF

64

摘要：

为了应对转向迟滞，提高无人矿卡横向控制的精度，提出了一种基于模糊模型预测控制（FMPC）的无人矿卡横向控制算法。建立车辆动力学模型和跟踪误差模型，设计基于动态预瞄时间的车辆状态预测方法，根据预瞄后的车辆状态计算跟踪误差。进一步，通过结合模糊控制，设计了基于横向误差与航向角误差自适应调节权重矩阵的模型预测控制（MPC）控制器。利用硬件在环仿真实验与实车实验验证所提算法的有效性。结果表明：硬件在环仿真实验中，FMPC算法的最大横向误差相较于纯追踪（Pure Pursuit）算法减少了43.0%；实车实验中，空载上山和重载泊入停车场2种工况下，FMPC算法相较于Pure Pursuit算法，最大横向误差分别减少了50.1%和17.6%。FMPC算法控制效果优于Pure Pursuit算法，显著提升了无人矿卡横向控制的精度。

关键词: 无人矿卡, 横向控制, 转向迟滞, 动态预瞄, 模糊模型预测控制（FMPC）

Abstract:

To address the issue of steering lag and improve the accuracy of lateral control in unmanned mining trucks, this study proposed a lateral control algorithm based on fuzzy model predictive control (FMPC). First, the vehicle dynamics model and tracking error model were established. Subsequently, a vehicle state prediction method based on dynamic preview time was designed, and the tracking error was calculated according to the predicted vehicle state after the preview period. Furthermore, by integrating fuzzy control with model predictive control (MPC), an MPC controller was developed that adaptively adjusts the weight matrices of both lateral error and heading angle error. The effectiveness of the proposed FMPC algorithm was validated through hardware-in-the-loop simulation experiments and real-vehicle tests. The results indicate that, in the hardware-in-the-loop simulation, the maximum lateral error of the FMPC algorithm is reduced by 43.0% compared to the Pure Pursuit algorithm. In real-vehicle experiments conducted under two operational conditions—empty-load uphill driving and heavy-load parking—the maximum lateral errors are reduced by 50.1% and 17.6%, respectively, in comparison to the Pure Pursuit algorithm, demonstrating that the FMPC algorithm achieves superior control performance and significantly enhances the lateral control accuracy of unmanned mining trucks.

Key words: unmanned mining trucks, lateral control, steering lag, dynamic preview, fuzzy model predictive control (FMPC)

中图分类号: