Research On SLAM And Path Planning Algorithm Of Autonomous Mobile Robot

With the continuous improvement of science and technology,more intelligent robots appear in people's vision.Autonomous mobile robot is a typical representative among them,which has been widely applied into in life,industry,medical and military,etc.With the increasing application of autonomous mobile robots,its surroundings are becoming more complex.How to effectively improve the abilities of environment perception and path planning of autonomous mobile robots under complex environments has become a top priority.This thesis deeply studies the key technologies Simultaneous Localization And Mapping(SLAM)and path planning of autonomous mobile robots,and the main research contents are as follows:(1)For the traditional EKF-SLAM algorithm,it is difficult to obtain accurate system noise and measurement noise in the actual application process which will cause the problem of low positioning precision.The fuzzy system is proposed to dynamically adjust the covariance matrix of the measurement noise,and the optimal parameters of the fuzzy system is solved by the improved particle swarm optimization algorithm,which could effectively improve the positioning precision of SLAM algorithm of the mobile robot.(2)Aiming at the particle degradation problem caused by the particle diversity reduction during the resampling process of RBPF-SLAM algorithm,an adaptive genetic resampling algorithm is proposed to solve the problem of particle degradation which can effectively improve the diversity of particles and the SLAM algorithm performance of mobile robot.(3)Aiming at the problems of low safety,many inflection points,and uneven trajectories in the traditional A* path planning algorithm,the efficiency and safety of global path planning under static environment are improved by the optimization of valuation function,search method,and generation trajectory which could make the planning path more in line with the movement of mobile robots.Aiming at the problem of difficult evaluation function coefficient selection of DWA local path algorithm,simulation verification of different evaluation function coefficients in DWA algorithm is done to provide a guidance for the practical application of DWA algorithm.(4)The autonomous mobile robot navigation system is designed based on the ROS robot operating system.Firstly,the improved adaptive genetic resampling RBPF-SLAM algorithm is verified through the map construction experiments under different environments.The experimental results show that the improved adaptive genetic resampling RBPF-SLAM algorithm can complete the accurate extraction of the surrounding environment features and realize the accurate map construction.Then,the effectiveness of the proposed path planning algorithm for mobile robots is verified through path planning experiments in static and dynamic environments.The improved path planning algorithm could well adapt to different environments.Finally,the algorithm precision of the autonomous mobile robot navigation system is tested.The experiment results show that the positioning precision of the autonomous mobile robot navigation system is within 7cm,which has a certain practical application value.