Research On SLAM Algorithm Of Mobile Robot Applied In Complex Environment

With the rapid development of artificial intelligence technology,robots have been widely used in industrial production,daily life,environmental detection and other fields.Simultaneous localization and mapping(SLAM)is the basis of autonomous mobility of robots.However,in complex environments,sensors carried by mobile robots will degrade the measurement accuracy due to external interference,thus reducing the accuracy of localization and mapping of robots.Therefore,this paper studies how to improve the accuracy of robot positioning and mapping in a complex environment,which is of great significance to accelerate the application of robot in the actual environment.The content of this paper is as follows:(1)This paper describes the problem of mobile robot SLAM,introduces the related research schemes of laser SLAM,elaborates the principle and framework of SLAM algorithm based on particle filtering,and analyzes the reason why the accuracy of robot localization mapping is decreased due to the limitation of particle filtering algorithm.(2)In order to reduce the accuracy of robot localization mapping due to the loss of particle diversity caused by resampling,brain storm optimization improved particle filtering SLAM(BSO-SLAM)algorithm was proposed.K-means clustering operation is completed according to the difference of particle weight,and cross-mutation processing is carried out for the set after clustering.The resampling of particle filter is replaced by brain storm optim ization,so as to alleviate the dilution of particles,increase the diversity of particles,and improve the localization accuracy of SLAM algorithm.The simulation results show that compared with GA-FastSLAM2.0 algorithm,BSO-SLAM algorithm improves the accuracy of location mapping,and the average positioning accuracy error of the algorithm is reduced by 24%.(3)For the problem of high time complexity of BSO-SLAM algorithm,the optimized SLAM algorithm based on plant cell swarm algorithm(PCSA-SLAM)is proposed.Plant cell swarm algorithm was used to adjust the particle distribution after the importance sampling of particle filter,so that the particle distribution was concentrated in the high likelihood region,and the resampling process was eliminated.Thus,the degradation of particle weight and the loss of particle diversity can be solved,and the filtering accuracy can be improved.The simulation results show that,compared with the GFA-FastSLAM2.0 algorithm,the PCSA-SLAM algorithm improves the mapping accuracy,the average positioning accuracy error of the algorithm is reduced by 26%,and the road sign prediction error is reduced by 45%.(4)A robot experimental platform was built,and PCSA-SLAM algorithm was applied to the localization mapping task of mobile robot.The experimental results show that under different environments,the PCSA-SLAM algorithm can effectively represent the contour of the actual terrain and the obstacle information in the environment,and the map constructed by it has higher accuracy and stronger robustness.