基于改进快速搜索随机树算法的包裹分拣路径规划算法

Improved RRT-based sorting path planning algorithm for parcel

针对现有快递包裹分拣系统存在的传输速度慢,准确率低等问题,提出一种改进的快速搜索随机树(RRT)算法。该算法以RRT算法为基础,首先建立了包裹环境模型,引入人工势场法引力分量使节点的扩展更具方向性。其次,对于节点进入障碍物区域需多次重新采样的问题,采用扇形区域法避障以提高算法生成质量,并在路径规划结束后采取二次优化,以获得平滑的路径。仿真结果表明,改进后的算法相比于传统的RRT算法,包裹传输速度提高73.22%,与加入人工势场法的改进算法相比,包裹传输速度提高94.46%。

To solve the problems of slow transmission speed and low accuracy in the existing express parcel sorting system,an improved Rapidly-exploring Random Trees(RRT)algorithm was proposed.Based on RRT algorithm,the wrapping environment model was first established,and the gravitational component of artificial potential field method was introduced to make the extension of nodes more directional.For the problem that nodes need to be resampled many times to enter the obstacle area,the sector area method was adopted to avoid obstacles to improve the quality of algorithm generation.After finishing the path planning,the secondary optimization was adopted to obtain a smooth path.The simulation results showed that the parcel transmission speed was increased by 73.22%compared with the traditional RRT algorithm,and the parcel transmission speed was increased by 94.46%compared with the improved algorithm of the artificial potential field method.