汽车工程

汽车工程 ›› 2024, Vol. 46 ›› Issue (4): 588-595.doi: 10.19562/j.chinasae.qcgc.2024.04.004

• • 上一篇    

基于复合动态采样的自动驾驶矿车节能路径规划方法

丁志杰1,王亚飞1,2(),章翼辰1,邬明宇1,王亦乐1   

  1. 1.上海交通大学机械与动力工程学院，上海 200240
    2.上海交通大学，汽车动力与智能控制国家工程研究中心，上海 200240
  • 收稿日期:2023-08-28 修回日期:2023-10-14 出版日期:2024-04-25 发布日期:2024-04-24
  • 通讯作者: 王亚飞 E-mail:wyfjlu@sjtu.edu.cn
  • 基金资助:
    国家自然科学基金(52072243)

Energy-Saving Planning Method for Autonomous Driving Mining Trucks Based on Composite Dynamic Sampling

Ting Chikit1,Yafei Wang1,2(),Yichen Zhang1,Mingyu Wu1,Yile Wang1   

  1. 1.The School of Mechanical Engineering，Shanghai Jiao Tong University，Shanghai 200240
    2.Shanghai Jiao Tong University，National Engineering Research Center for Automotive Power and Intelligent Control，Shanghai 200240
  • Received:2023-08-28 Revised:2023-10-14 Online:2024-04-25 Published:2024-04-24
  • Contact: Yafei Wang E-mail:wyfjlu@sjtu.edu.cn

RichHTML

2

PDF (PC)

42

赞

摘要：

近年来，考虑安全与效率的自动驾驶矿车路径规划方法已逐渐成熟，并在多种矿山场景落地应用。与此同时，产业界和学术界也开始关注如何利用路径规划提升矿车的燃油经济性。针对这一需求，本文提出了一种矿山场景下的自动驾驶矿车节能路径规划方法，其主要特点是根据车速、道路坡度及障碍物进行S-L（进度-偏离）和S-T（进度-时间）的复合动态采样。针对矿山典型地形场景，建立了矿车燃油消耗指标，提出了安全性-运行效率-能耗综合路径评价模型；为了防止评价模型的各项权重陷入局部最优，设计了基于模拟退火策略的粒子群自适应优化方法。在矿山实际场景的测试中，本研究提出方法较现有方法在燃油经济性指标上平均提升了11.28%。

关键词: 智能汽车, 自动驾驶, 燃油经济性, 局部路径规划

Abstract:

In recent years， path planning methods of autonomous driving mining truck that consider safety and efficiency have been gradually maturing and have been implemented in various mining scenarios. Simultaneously， the utilization of path planning methods to improve the fuel efficiency of mining trucks is paid more and more attention to by both the industry and academia. In response to this requirement， a method for energy-efficient path planning of autonomous driving trucks within mining environments is proposed in this paper. Its primary features encompass the utilization of composite dynamic sampling for S-L （Station - Lateral deviation） and S-T （Station - Time）， based on speed， road gradient， and obstacles. A fuel consumption index for typical terrain scenarios in mining environments is established. Additionally， a comprehensive path evaluation model of safety， efficiency and energy consumption is introduced. To prevent the entrapment of the evaluation model's weights in local optima， an adaptive optimization method based on the particle swarm algorithm with simulated annealing strategy is designed. Through testing in real mining scenarios， the method proposed in this paper has exhibited an average improvement of 11.28% in fuel economy metrics compared to existing methods.

Key words: intelligent vehicle, autonomous driving, fuel economy, local path planning