Research On Path Planning Algorithm Based On Obstacle Avoidance For Mobile Robots

Mobile robots have integrated sensor technology,electromechanical automation control,computer,artificial intelligence and other interdisciplinary technologies,whose development is an important manifestation of a country's high-tech level and industrial automation.In recent years,as one of the key technologies for real-world applications related to robotics,path planning based on obstacle avoidance has gradually become a hotspot and focus of research by scholars from all over the world.In the process of path planning,mobile robots need to plan an optimal path for obstacle avoidance from the starting point to the target point.Therefore,this topic has theoretical and practical significance for the research on path planning technology based on obstacle avoidance for robotics.This paper analyzes and studies path planning related to mobile robots,including global path planning in general environments,global path planning in raster map environment,and hybrid path planning containing local path planning.(1)For the global path planning part related to mobile robots,an improved cooperative quantized gray wolf algorithm is proposed.Seven types of international general test functions show that the improved collaborative quantized gray wolf algorithm has improved convergence speed,accuracy and stability.And it is successfully applied to the established circular obstacle environment model.At the same time,it also solves the problems,such as slow convergence speed and excessive search path cost in the traditional gray wolf algorithm while avoiding obstacles.(2)In view of the problem of unsmooth trajectory of mobile robots in path planning,a cost function constrained by the stability cost function is established.In addition,given the large redundancy problem caused by the improved collaborative quantized gray wolf algorithm after adding the clone and merge operation,an adaptive probability search gray wolf algorithm is proposed and it is applied to raster map environment composed of multiple obstacles,which not only solves large redundancy problem with regard to the algorithm itself,but also overcomes the problem,such as trajectory cost consumption and smoothness.(3)More importantly,aiming at the shortcomings of mobile robots in global path planning and local path planning,a hybrid path planning algorithm idea is proposed.The method first uses an improved adaptive probability search algorithm to search for an optimal path in a grid environment,then use the nodes of the trajectory as the sub-target points,and finally uses an improved artificial potential field method to complete local path planning,which not only solves the problem such as unreachable targets in the Artificial Potential Field method,but also is successfully applied to the hybrid path planning.(4)Last but not least,for the verification part of the mobile robot path planning platform,firstly,the trajectory verification with regard to the adaptive gray wolf algorithm in the grid map environment was completed under the ubuntu virtual experiment platform based on Linux,and then the verification containing the improved artificial potential field method about hybrid path planning was realized under the joint simulation of VREP and MATLAB.Eventually,a four-wheel mobile robot based on ROS physical tracking platform developed was built.On the basis of the theoretical research in this paper,the validity of the path planning algorithm is verified by means of the tracking physical platform.