智能汽车平行泊车路径规划算法  

Parallel parking path planning algorithm for intelligent vehicles

作　　者：武继权 杜峰[1,2] 程剑锋 徐慧 陈强 WU Jiquan;DU Feng;CHENG Jianfeng;XU Hui;CHEN Qiang(School of Automotive and Transportation,Tianjin University of Technology and Education,Tianjin 300222,China;National and Local Joint Engineering Center for Intelligent Vehicle-Road Collaboration and Safety,Tianjin 300222,China)

机构地区：[1]天津职业技术师范大学汽车与交通学院,天津300222 [2]智能车路协同与安全技术国家地方联合工程研究中心,天津300222

出　　处：《天津职业技术师范大学学报》2025年第1期39-46,共8页Journal of Tianjin University of Technology and Education

基　　金：中国高校产学研创新基金项目(2022IT178);天津市研究生科研创新项目(2022SKYZ388).

摘　　要：针对自动泊车路径规划时在平行泊车过程中存在曲率突变和曲率变化波动的问题,提出一种基于改进蜣螂算法的平行泊车路径规划算法。该方法基于圆弧-直线-圆弧路径规划原则,计算平行泊车的避障约束,构建一个基于五次多项式曲线的非线性路径规划模型,并纳入多种约束条件,最终通过优化的蜣螂算法解决这一多约束问题。仿真和实车试验结果表明:优化的蜣螂算法相比原始算法减少约54.5%的迭代次数,有效提升收敛速度和寻优性能;获得的平行泊车路径长度相较于传统方法缩短3.48%;在不同条件下,路径规划均展现出良好的跟踪性能,验证了该泊车路径规划算法的可行性与可靠性。To address issues of curvature discontinuity and fluctuation during parallel parking in parallel automatic parking path planning,this paper proposes an improved dung beetle optimization(DBO)algorithm-based parallel parking path planning algorithm.This method applies the arc-line-arc path planning principle to calculate the obstacle-avoidance constraints for parallel parking.A nonlinear path planning model based on quintic polynomial curves is constructed,incorporating multiple constraints,and finally the optimized DBO is used to solve this multi-constraint problem.Simulation and real vehicle test results show that the optimized DBO reduces the iteration count by approximately 54.5%compared to the original algorithm,effectively enhancing convergence speed and optimization performance.The length of the parallel parking path is reduced by 3.48%compared to traditional methods.Under various conditions,the path planning demonstrates excellent tracking performance,confirming the feasibility and reliability of the proposed parking path planning algorithm.

关 键 词：驾驶辅助 平行泊车 路径规划 五次多项式 蜣螂优化算法 

分 类 号：U464.6[机械工程—车辆工程]