In recent years,due to the development of scientific and technological progress,various types of robots have begun to enter various fields to help complete various tasks,such as various surveying robots in hazardous environments,and palletizing robots in industrial environments.The vigorous development of the robot business has improved people's living environment,playing an increasingly important role in human daily life.Simultaneous localization and mapping(SLAM)technology and path planning(Path Planning),as the key to realize the autonomous navigation capability of mobile robots,have very profound research significance.In this paper,based on the two-dimensional lidar as the robot indoor observation system,the SLAM technology and path planning technology are theoretically studied and simulated experiments,and a low-cost lidar-based SLAM and robot operating system(ROS)are designed.Path planning system,which can complete the construction with high accuracy,and meet the obstacle avoidance and path planning in the indoor environment.Firstly,the kinematics model and environment map model established in this paper are introduced,two kinds of lidar mathematical models for lidar are compared and analyzed,and the likelihood field model is selected as the lidar ranging model in this paper.Analyze the coordinate conversion and the construction algorithm of raster map.Then,the theoretical research on SLAM based on particle filtering is mainly carried out.The SLAM method based on Quantum Particle Swarm Optimization(QPSO)proposed in this paper is theoretically deduced,analyzed by MATLAB simulation,and compared with the traditional particle filtering method,FAST_SLAM for pose and feature point errors.The superiority of the proposed algorithm is proved,and the laser SLAM loopback detection technology is studied,and a delay decision algorithm is proposed to improve the robot loopback detection technology.Secondly,the robot path planning algorithm is mainly studied,which is divided into two parts: global path planning and local path planning.The traditional global path planning algorithm Dijkstra algorithm and A* algorithm are compared and analyzed,and the priority search strategy based on the two methods is analyzed through MATLAB simulation,and the path planning process based on genetic algorithm is deduced and the parameters are analyzed through MATLAB The impact on the smoothness of the path,and finally the dynamic window method(DWA)local path planning algorithm is deduced,and the results are simulated.Finally,the overall framework of building a robot is explained from two aspects of hardware and software,and the built robot is used as a test platform to actually test the robot's SLAM and path planning functions.The QPSO-SLAM method proposed in this paper is transplanted to the robot platform,and compared with the traditional two particle filter-based SLAM methods,which proves the robustness and efficiency of the algorithm in this paper.Then test the path planning function of the robot,view the obstacle avoidance and driving situation of the robot in real time through the RVIZ tool,and analyze all the test results. |