Research On Autonomous Mapping And Navigation Technology Of Tracked Robot In Unknown Environment Based On ROS

Robots are widely used in industry and life services.The country is in the era of old house demolition and transformation.It is a challenging task to explore the surrounding unknown environment in advance during the transformation process.Real-time location mapping and navigation are two of them.Key point.Aiming at the problems of poor software reuse rate and weak environmental adaptability of the robot,the robot designed in this project is equipped with ROS(Robot Operating System)system,and the walking mechanism adopts crawler type.The system software of the robot is designed and developed with modular nodes,and the lidar data nodes,keyboard operation nodes,chassis control nodes,simultaneous positioning and mapping nodes,path planning and navigation nodes,and other auxiliary nodes are given.Aiming at the kinematics of the crawler robot,the kinematics analysis of the crawler robot is performed.The crawler robot chassis drive mode is designed as a two-wheel differential drive.The robot is controlled to move flexibly to the corresponding speed by using the speed difference between the two drive wheels of the robot chassis.Target point.Incremental photoelectric encoders are attached to the two drive wheel motors.Through displacement increments,odometry information representing the relative position coordinates and heading angle of the robot chassis is calculated,that is,the control information of the robot chassis at the current moment.Aiming at the problem of positioning and mapping of tracked robots,an improved resampling Rao-Blackwellized particle filter for simultaneous positioning and map construction(RBPF-SLAM)algorithm was proposed to improve the inaccuracy of grid map construction caused by particle weight degradation and particle scarcity problems.The feasibility and accuracy of the improved algorithm is verified by simulation experiments.The multi-sensor based information fusion positioning is used to overcome the positioning errors caused by chassis sliding,thereby improving the positioning accuracy.Simulations in MATLAB environment and positioning experiments in Gazebo environment are given to verify the effectiveness of the improved algorithm.Aiming at the problems of "U" type confusion in the process of local path planning,a DWA algorithm based on global paths was designed to improve the evaluation function of local path planning.Simulation experiments verify the feasibility and accuracy of the DWA algorithm based on the global path.Import the map built by the tracked robot through the improved algorithm into the corresponding code package,write the navigation function package and the corresponding Launch startup file,and perform navigation.Through the analysis of relevant data,it is shown that the path planning algorithm in this chapter has better accuracy than the traditional algorithm.And stronger robustness.