基于RBF神经网络的自动泊车路径规划

文章通过逆向路径规划分析平行泊车过程的可能碰撞点和计算泊车所需的最小泊车空间,用泊车初始区域代替传统路径规划的初始点,实车试验采集泊车过程的数据,采用不同的数据样本用于粒子群优化的RBF神经网络,避免对安全距离等多种约束关系的分析,使规划的泊车路径能较好适用于实际泊车过程。仿真结果和实车试验均表明按照上述方法生成的路径泊车成功率较高。

Path planning of automatic parallel parking based on RBF neural network

In this paper,based on the reverse path planning,the possible collision points in the process of parallel parking are analyzed and the minimum parking space needed is calculated.To get a better planned parking path applied to actual parking process and avoid the analysis of a variety of constraint relations such as safe distance,the initial point of the traditional path planning is replaced with the initial region,the data during the whole process of parking is collected by real vehicle tests,the different data samples are used for training radial basis function(RBF) neural network by particle swarm optimization to generate the parking path.Simulation results and real vehicle tests show that the proposed method achieves a higher parking success rate according to the path generated with the above method.