Research On AGV Navigation And Path Tracking Control Based On Laser SLAM

With the increasing improvement of scientific and technological level,more and more intelligent robots are invested in various fields to help human beings complete various tasks.Automatic guided vehicle(AGV)integrates the functions of automatic navigation,automatic control and real-time communication.AGV has the characteristics of high flexibility,high automation and high work efficiency.It is widely used in warehousing,logistics,intelligent inspection and so on.With the popularization and application of intelligent robots more widely,automatic driving technology is also more critical.Therefore,how to accurately position AGV in complex environment and make reasonable planning and path planning has become a key research content.This paper mainly focuses on the real-time positioning,map construction,path planning and obstacle avoidance control of laser guided AGV.The main contents include:Firstly,this paper introduces the framework and operation structure of robot software development platform ROS,analyzes the construction process of simulation environment,establishes the kinematics model of AGV and the simulation model of sensor,and uses grid map to build the simulation environment through ROS system.Secondly,aiming at the problem of large measurement error of single sensor,combined with the fusion positioning method of Monte Carlo particle filter and extended Kalman filter,it is used to improve the global positioning accuracy of AGV.Aiming at the global path planning problem of AGV,based on the traditional A* algorithm,the optimization method of quantitative environment and eliminating redundant points are designed to optimize the calculation time and path length of the algorithm.Then,aiming at the local path planning problem of laser guided AGV,obstacle avoidance and target tracking control of AGV are realized based on nonlinear model predictive control theory.The improved scheme of artificial potential field method is introduced into the model predictive controller to improve the control effect of the controller in a complex environment.Finally,the actual vehicle test platform of laser guided AGV is built,the positioning and mapping of laser guided AGV are carried out in the actual environment.At the same time,an improved global path planning method is added through the ROS plug-in,and the MPC optimization program is converted into C code,and the validity of the control algorithm is verified through the built test platform.