Research On The SLAM And Path Planning For Mobile Robots

With the rapid development and continuous progress of science and technology in modern society,the development of robots has become an important sign of technological progress.Now,the vigorous development of artificial intelligence technology has brought robots into a new era.In order to realize how mobile robots clearly understand how to react and move in an unknown environment,how to use sensors to sense the transformation of the external environment and transfer information to the controller,positioning,map building and navigation of the robot have very important research significance and development prospect.This paper mainly studies the two types of problems of sIMUltaneous localization and mapping(SLAM)and path planning of mobile robots.First,it introduces the origin of robots,summarizes the research achievements of robot products at home and abroad,and the main application fields of robots in the development process.Introduce the filters commonly used when mobile robots do SLAM and the main technologies currently used by mobile robots in path planning.Secondly,the basic theoretical overview of the robot SLAM problem and the mathematical expressions of the robot state model,prediction model and observation model used in SLAM are introduced.The principle and process of Kalman filtering and extended Kalman filtering are introduced,and the SLAM algorithm of the extended Kalman filter is used in MATLAB to sIMUlate the robot's own SLAM to verify the effectiveness of the algorithm and analyze the results.Then,the path planning of mobile robots is studied.It mainly introduces the path planning algorithm based on the grid map: Dijistra algorithm and A * algorithm,respectively using Dijistra algorithm and A * algorithm for sIMUlation experiments;also based on the grid map,a robot path planning algorithm based on sIMUlated annealing optimization is proposed In this algorithm,the coordinates of the intermediate point are used as optimization variables,so that the robot can clarify the direction of movement when moving in the sub-path between the coordinates of the intermediate point,and the path length is the target function,so that the robot can plan a shorter path;finally The path planning algorithm of the traditional artificial potential field method was studied,the influence of the parameters of the traditional artificial potential field on the normalized path was analyzed,and the study on the optimization of the parameters using differential evolution algorithm was proposed.SIMUlation experiments were conducted in MTLAB to verify the algorithm The effectiveness.Finally,the ROS robot operating system is introduced.The positioning and mapping experiments of the Gmapping algorithm and the Hector algorithm are performed under the ROS system,and the mapping effects of the two algorithms are analyzed.The sIMUlation experiments of amcl positioning and move_base autonomous navigation under the ROS system.