Research On Indoor Navigation Of ROS Wheeled Robots Based On Laser SLAM Algorithm

With the continuous development of artificial intelligence,more and more wheeled robots with autonomous navigation functions have been applied to various fields.How the wheeled robot realizes its own positioning,constructs the environment map and path planning in the indoor complex environment has always been the focus of research.Therefore,this paper primarily focuses on investigating the core technology of navigation,namely Simultaneous Localization and Mapping(SLAM),as well as path planning.It aims to propose a navigation scheme based on laser RBPF-SLAM and the fusion of A* and DWA path planning algorithms.The key research objectives of this study are as follows:In response to the challenges of low positioning accuracy and insufficient particle diversity in the RBPF-SLAM algorithm,an enhanced laser SLAM algorithm is proposed.Correct the odometry information by using the lidar data to improve the initial pose accuracy.The observation model information and motion model information are fused to optimize the proposed distribution and reduce the error between the proposed distribution and the target distribution.The particle swarm optimization(PSO)algorithm is used to adjust the sampled particle set,so that the particles move to the area with high likelihood and improve the accuracy of the particle estimation.In the improved resampling algorithm,an adaptive local linear resampling algorithm is proposed,which uses the linear resampling method to sample highweight and low-weight particles and preserve particle diversity.The improved RBPF algorithm is simulated and compared.The test shows that the improved algorithm can effectively improve the estimation accuracy of particles and alleviate the degree of particle degradation.In the Gazebo environment,the improved RBPF-SLAM algorithm is simulated for building maps.The results show that the improved RBPF-SLAM algorithm is superior to the original algorithm in terms of running time,positioning accuracy and map details.Aiming at the problems of long search time and many turning points in A* algorithm for global path planning,an improved algorithm is proposed.The dynamic weighting coefficient of the ratio of the number of grids in the node area is added to the heuristic function,and the heuristic function is optimized by using the minimum cost value node selection strategy to improve the global search efficiency of the algorithm,and the collinear key point deletion strategy is used to smooth the global path.Integrate and improve A* and DWA as the robot path planning algorithm,and add the key point information of the global path to the DWA algorithm to ensure that the robot trajectory is close to the global optimum.The simulation test shows that the search time and the number of turning points of the improved A* algorithm are smaller than the original algorithm,and the path is smoother.The fusion algorithm can plan a global optimal path without collision.Using the two-wheeled differential wheeled robot equipped with the Robot Operation System(Robot Operation System,ROS)as a platform,the map construction and autonomous navigation experiments were carried out in different indoor environments.The experiments showed that the wheeled robot navigation based on the laser SLAM algorithm designed in this paper The system has better robustness.