Research On Path Planning Technology Of Wheeled Robot Based On ROS

Autonomous navigation is an important component of robot functions in the field of indoor service robot technology.And path planning is the technical basis for the robot to complete various functions in the application scenario of autonomous navigation technology.Algorithms of global and local path planning technology still have the phenomenon that the planning path is not smooth and the stability is not good.Aiming at the above problems,this paper takes wheeled robots as the research object and studies the path planning technology.Based on the ROS(Robot Operating System)platform,the framework of the autonomous navigation system is completed.To ensure the smooth progress of simulation and experiment,the kinematics model and lidar observation model of the wheeled robot are derived,and the simulation model of the wheeled robot is established based on this,in order to carry out the simulation and experimental verification of the path planning algorithm conveniently.On the global path planning algorithm,the A* algorithm is used to obtain the optimal path in the global scope.Aiming at the problem that the path planned by the A* algorithm is closer to the obstacle,by using the method of moving grid nodes near the obstacle out of the path search range,the improved path can maintain a safe distance from the obstacle.Aiming at the problem of unsmooth path,the phenomenon of excessively large path turning angles are reduced effectively by the method of increasing the number of path points and setting up the angle deviation threshold of adjacent path points,so the path become smoother.The simulation results show that the improved A* algorithm can avoid obstacles in both simple and complex scenarios and can control the turning angle of the path within 45 degrees.On the local path planning algorithm,the timed elastic band(TEB)algorithm is used to optimize the local planning path.To improve the stability of the TEB planning algorithm,it is integrated with the PID control algorithm to stabilize the speed and angular velocity commands.The fluctuation of is reduced effectively and the unstable phenomenon of control commands is avoided in the improved algorithm.The simulation results show that in a simple scenario,the variances of velocity and angular velocity commands are reduced by 84.2% and 46.2%,respectively,and in complex scenarios,the variances of velocity and angular velocity are reduced by 44.4% and 38.4%,respectively.A multi-task navigation scheduling algorithm is proposed to meet the needs of multi-target task planning for wheeled robots.The algorithm can determine the number of targets and specific locations by human-computer interaction,and it makes the distance between the robot and the target point as the principle of multi-task navigation scheduling.Finally,the algorithm realizes the scheduling function of near first and farther.The simulation results show that the multi-task navigation scheduling algorithm can navigate according to the principle of near-tofar in both simple and complex scenarios and stop the robot within a preset range of 0.2m from the target point.The results of experiments show that the global path is smoothed effectively,the phenomenon of crossing obstacles is circumvented and the turning angle are controlled within45 degrees successfully in the improved global algorithm.The fluctuations of control command are reduced in improved local planning algorithm.In simple scenarios,the velocity and angular velocity variances are reduced by 40% and 14.3%,respectively.And in complex scenarios,the velocity and angular velocity variances are reduced by 47.1% and 18.2%,respectively.The multi-task navigation scheduling algorithm can complete multi-task scheduling according to the principle of near-to-far,and make the robot stop within a preset range of 0.2m from the target point.