SLAM And Path Planning Methods For Lidar-based Mobile Robots

With the continuous development of robotics-related technologies,the demand for intelligent robots in various industries has increased,and SLAM technology and path planning are the technical keys to realize the autonomy of mobile robots,which is an important part of the Made in China 2025 that focuses on the development of intelligent mobile robot industry.Therefore,the research work in this paper has far-reaching significance as well as practical application value for the robot to realize industrialization and intelligence.The main research works in this paper are as follows.Firstly,the general framework of mobile robot is formulated,and the mecanum wheeled mobile robot with lidar is formed based on the mobile chassis built by the group independently in the early stage,and the models of each system of the mobile robot and the map model are introduced.Next,the principle knowledge of SLAM technology is briefly described,and the Hector SLAM algorithm and the particle filter-based Gmapping algorithm are studied and the corresponding simulation and map building experiments are carried out.The Gmapping algorithm based on fused data is proposed to address the accumulated odometer error caused by wheel slip in the particle filter-based Gmapping algorithm,which is a new hybrid proposed distribution by fusing the odometer information with IMU data and then combining the observation information to achieve a more accurate target distribution.The simulation world is built under Gazebo platform and the corresponding simulation experiments are conducted.The experimental results show that the improved Gmapping algorithm can effectively reduce the particle degradation problem and reduce the negative impact of the accumulated odometer error caused by wheel slip on the map building accuracy.Again,the path planning algorithm of the mobile robot is analyzed,including the global path planning algorithm and the local path planning algorithm.Among them,the A* algorithm in the global path planning algorithm is optimized,and a directional search-based A* algorithm and a Bessel curve smoothing path planning algorithm are proposed.By adding the key point extraction strategy and "straightening" processing strategy,the path redundancy points are removed and the number of path turns is reduced;in addition,the Bessel curve is introduced to smooth the spikes and inflection points in the path,which can better meet the actual needs of mobile robot path planning.Finally,experiments related to SLAM and path planning algorithms for mobile robots are conducted in an indoor environment using the lidar-based mecanum wheeled mobile robot experimental platform.The experimental results demonstrate the feasibility of the proposed theory and method.