汽车工程 ›› 2023, Vol. 45 ›› Issue (12): 2299-2309.doi: 10.19562/j.chinasae.qcgc.2023.12.012

所属专题: 智能网联汽车技术专题-规划&决策2023年

• • 上一篇    下一篇

基于双层求解策略的平行泊车轨迹规划

张洪昌1,2,3,宁鹏1,2,3,杨杰1,2,3,宋建伟1,2,3,郝麟1,2,3,曾娟1,2,3()   

  1. 1.武汉理工大学,现代汽车零部件技术湖北省重点实验室,武汉 430070
    2.武汉理工大学,汽车零部件技术湖北省协同创新中心,武汉 430070
    3.武汉理工大学,湖北省新能源与智能网联车工程技术研究中心,武汉 430070
  • 收稿日期:2023-04-05 修回日期:2023-05-11 出版日期:2023-12-25 发布日期:2023-12-21
  • 通讯作者: 曾娟 E-mail:zhc112@126.com
  • 基金资助:
    教育部创新团队发展计划(IRT_17R83);新能源汽车科学与关键技术学科创新引智基地(B17034);武汉理工大学重庆研究院科技创新研发项目(YF2021-07)

Parallel Parking Trajectory Planning Based on Double-Layer Solution Strategy

Hongchang Zhang1,2,3,Peng Ning1,2,3,Jie Yang1,2,3,Jianwei Song1,2,3,Lin Hao1,2,3,Juan Zeng1,2,3()   

  1. 1.Wuhan University of Technology,Hubei Key Laboratory of Modern Auto Parts Technology,Wuhan 430070
    2.Wuhan University of Technology,Auto Parts Technology Hubei Collaborative Innovation Center,Wuhan 430070
    3.Wuhan University of Technology,Hubei Technology Research Center of New Energy and Intelligent Connected Vehicle Engineering,Wuhan 430070
  • Received:2023-04-05 Revised:2023-05-11 Online:2023-12-25 Published:2023-12-21
  • Contact: Juan Zeng E-mail:zhc112@126.com

摘要:

轨迹规划在泊车系统中承接了上层感知和下层控制,是缩短泊车时间和降低跟踪难度的关键手段。针对平行泊车轨迹规划难以兼顾轨迹的生成质量、泛化能力和计算效率等问题,提出了基于双层求解策略的平行泊车轨迹规划方法。该方法分为两层:第一层将平行泊车的路径分为由锚点连接的两段路径,采用路径倒推的思路寻找出锚点,分别规划从终点到锚点的路径和从锚点到起点的路径,之后对路径附加“时间最优”剖面,逆序得到特定时刻的状态量与控制量;第二层采用同步联立正交配置法,将平行泊车最优控制中连续的状态量和控制量转化为轨迹非线性规划的离散点,把第一层得到的状态量与控制量作为初始值带入非线性规划求得数值解。同时,建立了5种平行车位泊车场景模型并进行了仿真分析,结果表明,针对不同的泊车起始位姿和车位尺寸,均能规划出满足约束条件的最优轨迹,提高了轨迹的生成质量和泛化能力,并具有较好的计算效率。

关键词: 轨迹规划, 双层求解策略, 平行泊车, 数值优化

Abstract:

Trajectory planning plays a key role in shortening parking time and reducing tracking difficulty, as it connects the upper-level perception and the lower-level control in parking systems. However, it is challenging to balance trajectory quality, generalization ability, and computational efficiency in parallel parking trajectory planning. To address this issue, the parallel parking trajectory planning based on double-layer solution strategy (DLSS) is proposed. The strategy includes two layers: in the first layer, the parallel parking path is divided into two segments connected by anchor points. The anchor point is identified through a backward path planning approach. The paths from the endpoint to the anchor point and from the anchor point to the starting point are planned separately. Then, a "time-optimal" profile is added to the path, and the state and control variables at specific time points are obtained in reverse order. In the second layer, the simultaneous orthogonal configuration method is used to transform the continuous state and control variables in the parallel parking optimal control into discrete variables in the trajectory nonlinear programming. The state and control variables obtained in the first layer are used as the initial values for the nonlinear programming to obtain numerical solutions. Five parallel parking scene models are established and simulated, and the results show that the optimal trajectory that meets the constraints requirements can be planned for different parking starting postures and parking space sizes, which improves the generation quality and generalization ability of trajectories, and has satisfactory computational efficiency.

Key words: trajectory planning, double-layer solution strategy, parallel parking, numerical optimization