基于粒子群算法改进的轮式收获机纯追踪模型的路径跟踪研究

为满足轮式收获机地头收获路径跟踪精度要求，本研究提出了一种基于粒子群改进的带有预测特性的纯追踪路径跟踪算法。建立了轮式收获机运动学模型，推导了基于轮式收获机运动学模型的纯追踪路径跟踪算法。以收获机航向误差和横向误差为基础，构建了带有预测特性的隶属度函数，采用权重系数自适应方法，通过粒子群优化（PSO）算法，实现了实时动态确定最优前视距离。以玉米收获机为试验平台，开展了直线路径跟踪路面试验与“8”字曲线路径跟踪路面实验，试验结果表明：在1.5m/s速度时，直线路径跟踪的最大横向误差为4.39cm，最大航向误差为2.31°。在1m/s时，曲线路径跟踪的最大横向误差为5.24cm，最大航向误差为2.41°。试验结果表明本文设计改进的路径跟踪算法对直线路径及曲线路径都具有良好的路径跟踪效果，满足轮式收获机田间作业要求。

Research on path tracking of improved pure pursuit model of wheeled harvester based on PSO algorithm

In order to improve the path tracking accuracy of wheeled harvester, a pure pursuit path tracking algorithm with prediction characteristics based on particle swarm improvement is proposed. A kinematic model of the wheeled harvester is established, and a pure pursuit path tracking algorithm based on the kinematic model of the wheeled harvester is derived. Based on the harvester heading error and lateral error, an affiliation function with prediction characteristics is constructed, and the optimal forward-looking distance is determined in real time by the particle swarm optimization (PSO) algorithm using adaptive weighting coefficients. The test results show that the maximum lateral error of straight-line path tracking is less than 4.39cm and the maximum heading error is less than 2.31° under 1.5m/s speed condition and 1m/s condition. The maximum lateral error of curve path tracking is less than 5.24cm, and the maximum heading error is less than 2.41°. The test results show that the improved path algorithm designed in this paper has good path tracking performance for both linear path and curve path, and meets the requirements of corn harvester field operation.