基于新型CF-RRT^(*)算法的智能汽车新型自动泊车路径规划技术  

The New Automatic Parking Path Planning Technology for Intelligent Vehicles Based on a Novel CF-RRT^(*)Algorithm

作　　者：赵闵清 宁善平 ZHAO Minqing;NING Shanping(Product Development Technology Center,Jiangling Motor Co.,Ltd.,Nanchang 330000,Jiangxi,China;Guangdong Communication Polytechnic,Guangzhou 510650,Guangdong,China)

机构地区：[1]江铃汽车股份有限公司产品开发技术中心,江西南昌330000 [2]广东交通职业技术学院,广东广州510650

出　　处：《广东交通职业技术学院学报》2025年第2期59-63,共5页Journal of Guangdong Communication Polytechnic

基　　金：广东省教育厅青年创新人才项目(编号:2022KQNCX191,2021KQNCX177)。

摘　　要：针对原始RRT^(*)自动泊车路径规划算法难以获取平滑、准确及最优路径的问题,提出一种基于新型CF-RRT^(*)算法的自动泊车路径规划算法。首先,为提升避障安全性,对环境障碍物进行尺寸膨胀处理,并根据膨胀后的安全距离进行步长自适应调整;然后,为提升路径规划收敛性能,引入自适应权重修剪策略对冗余节点进行修剪;最后,引入三次样条曲线对路径进行平滑处理。仿真试验结果表明:与原始RRT^(*)算法相比,改进后的新型CF-RRT^(*)算法规划出的路径在两种自动泊车场景中不仅可有效实现避障泊车功能,而且规划时间和路径质量更具有优越性,可为泊车路径规划技术的发展提供理论依据和技术参考。In response to the difficulty of the original RRT^(*)algorithm in obtaining smooth,accurate,and optimal paths for autonomous parking,this study proposes an automatic parking path planning algorithm based on a novel CF-RRT^(*)algorithm.Firstly,to enhance obstacle avoidance safety,the environment obstacles are subjected to size expansion,and the step size is adaptively adjusted based on the expanded safety distance.Secondly,to improve the convergence performance of path planning,an adaptive weight pruning strategy is introduced to prune redundant nodes.Finally,a cubic spline curve is used to smooth the path.Simulation results demonstrate that compared to the original RRT^(*)algorithm,the improved CF-RRT^(*)algorithm not only effectively achieves obstacle avoidance parking functionality in two automatic parking scenarios but also exhibits superior planning time and path quality.This study can provide theoretical basis and technical references for the development of parking path planning technology.

关 键 词：自动泊车 路径优化 RRT^(*)算法 三次样条曲线 仿真试验 

分 类 号：U463.6[机械工程—车辆工程]