单舵轮AGV叉车的高精度导航算法研究与应用

Research and Application of High Precision Navigation

摘要:

针对单舵轮自动导引车(AGV)面向印刷车间复杂的作业环境下传统导航方法定位精度差和路径规划算法生成轨迹不固定的问题,研究并设计了完整环境地图下的固定路径和提高导航与定位精度的算法。针对自适应蒙特卡洛定位算法定位精度达不到作业要求,采取了改进的Cartographer-SLAM算法与PID算法结合控制循迹的方法,大大提高了定位导航精度；针对常规导航框架下的路径规划算法会由于AGV车避障行为而改变局部路径,从而改变导航过程中的全局路径,导致在狭窄车间环境下会发生碰撞的问题,设计了生成导航路径点的算法,生成由直线和弧线轨迹组成的固定路径,以用于AGV叉车的PID控制循迹,实现了固定路径的规划；实践证明,该方法解决了单舵轮AGV叉车在车间环境下采用传统导航方法出现的问题,提高了安全性和可控性,工作效率也随之提升。

Abstract:

Aiming at the problems of poor positioning accuracy of traditional navigation methods and unfixed path generated by path planning algorithm for AGV forklift with single steering wheel facing the complex working environment of printing shop, the fixed path under the complete environment map and the algorithm to improve navigation and positioning accuracy are studied and designed. As the positioning accuracy of single amcl (adaptive monte carlo positioning) algorithm can not meet the requirements of the operation, the higher precision cartographer positioning algorithm and PID algorithm are combined to control the tracking, which greatly improves the positioning and navigation accuracy. For the problem that the path planning algorithm under the conventional navigation framework will change the local path due to the obstacle avoidance behavior of AGV vehicles, thus changing the global path in the navigation process, resulting in collisions in a narrow workshop environment, a method combining Dijkstra algorithm to mark feature point composition is proposed to generate a fixed path consisting of straight lines and arc tracks for the PID control tracking of AGV forklifts, The fixed path planning is realized. The practice shows that this method solves the problems of traditional navigation method for AGV forklift with single steering wheel in the workshop environment, improves the safety and controllability, and improves the work efficiency.