| In recent years,robots have been widely used in people's lives.Mobile robots get rid of the shackles of wire harness and can help people to complete more complex work more flexibly.With the rapid development of robot technology,the traditional mobile robot has been unable to meet the requirements of all aspects of life.Therefore,the research of mobile robot autonomous navigation system based on ROS(Robot Operating System)operat-ing system mainly includes the framework design of robot navigation system based on ROS,robot positioning,map construction and global path planning.The main contents of this paper are as followsThe overall scheme design of robot navigation system.The ROS operating system and its advantages and disadvantages are simply studied and analyzed,and the software and hardware framework of the mobile robot navigation system is established.In this paper,the working principle,model selection and data acquisition of the main sensors: Lidar and IMU(Inertial Measurement Unit)are described in detail.SLAM algorithm analysis and sensor information fusion research.Google cartographer algorithm with closed-loop detection is selected as Slam(Simultaneous Localization and Mapping)mapping algorithm.In order to solve the problem of large positioning deviation when using wheel odometer alone,extended Kalman Filter is used to fuse odometer and IMU data to improve robot positioning accuracy.Simulation results show that the localization effect of mobile robot before and after sensor data fusion is compared.Research on path planning algorithm.Three common path planning algorithms are compared and analyzed.Aiming at the problems of A* path planning algorithm,an improved scheme is proposed.The main embodiment is: adding expansion factor to the original evaluation function of A* algorithm to increase the movement cost of sub nodes around obstacles;In order to reduce the redundancy of nodes,in all the optimal nodes generated by A* algorithm,if there is no obstacle between the front and back nodes,all the child nodes in the two nodes are deleted;Modify Open?Set set to reduce the number of traversal comparison.The simulation results show that the path planning effect of A*algorithm before and after the improvement is compared.Mobile robot navigation system experiment.Through the Mir100 experimental platform for experimental verification,in the indoor complex environment for mapping and navigation,the fixed starting point and target point repeated 12 experiments,and record and analyze the experimental data,the experimental results show that the optimization algorithm can more efficiently complete the task of mapping and path planning. |
PDF Full Text Request