Research And Application Of SLAM And Path Planning Algorithms Based On Mobile Robots

With the continuous development of mobile robot technology,its importance in various practical applications is becoming increasingly significant.As an autonomous mobile system,mobile robots need to perfectly coordinate perception,decision-making,and execution to complete autonomous tasks.Among them,environmental perception is the core of mobile robots,while Simultaneous Localization and Mapping(SLAM)and path planning are important components of environmental perception,and are also the foundation of autonomous navigation for mobile robot.Therefore,in this paper,SLAM technology and path planning algorithms for mobile robots based on Robot Operation System(ROS)are deeply conducted.The main research contents are as follows:(1)Proposed a Gmapping algorithm based on improved Rao-Blackwellized Particle Filter(RBPF)to address the problem of insufficient number of particles.The algorithm uses an adaptive number of particles method to dynamically adjust the particles count based on the current state and observation values.Meanwhile,the map is modeled as a grid map,allowing for incremental updates of the map during robot movement,reducing computation and time complexity.In contrast,the RBPF-SLAM algorithm requires re-estimation of the entire map after each measurement.The improved algorithm not only improves the performance and efficiency of the SLAM algorithm but also is suitable for various practical applications.(2)On the basis of the traditional A* algorithm,an improved A* algorithm is proposed in this paper.Firstly,a method for optimizing the heuristic function is proposed,which takes into account the influence of the parent node of the current node on the search path of its child nodes,thus reducing the computational cost of nodes.Meanwhile,the density of environmental obstacles is represented by introducing the obstacle ratio,which is then incorporated into the optimized heuristic function.This makes the mobile robot more flexible when encountering obstacles.Secondly,the improved A* algorithm is integrated with the Dynamic Window Approach(DWA),so that the planning results of the mobile robot are closer to the global path.Multiple sets of simulation experiments have verified that the improved A* algorithm exhibits significant advantages in path cost,search time,and number of traversed nodes.Compared with the traditional A* algorithm,the improved algorithm effectively solves its shortcomings.(3)An experiment was conducted to use the ROS system for SLAM mapping and path planning of mobile robots in real-world environments.Firstly,the EKF-SLAM,RBPF-SLAM,and an improved RBPF Gmapping algorithm were used to construct maps of the real environment,and the three algorithms were compared.The results showed that the improved algorithm was more accurate than the first two algorithms in constructing maps,and could clearly display the shape and position of obstacles.Secondly,actual path planning experiments were conducted on the original A*algorithm and the improved A* algorithm in two different obstacle environments,and the feasibility of the improved algorithm was verified.Finally,a robot navigation experiment was conducted,and the results showed that the fusion algorithm could avoid unknown obstacles in real-time and successfully guide the mobile robot to the destination.