Research On The Navigation Technologies Of Indoor Mobile Robot Based On ROS

With the development of intelligent robot,mobile robot is gradually developing from manual control to intelligent autonomous navigation.Autonomous navigation mainly includes map-building,localization,path planning and other key technologies.In this paper,some key technologies of autonomous navigation for mobile robots are studied and verified experimentally.Aiming at the problem of self-localization of mobile robots,the Kalman filtering is used to fuse the dead reckoning and inertial position,and combined with Monte Carlo positioning to realize mobile robot self-localization.The raster map is constructed by using LIDAR data and the concept of grid occupancy is introduced to reduce the influence of sensor noise.The motion model is predicted by a circle with a radius of 1.2v centered on the current mobile robot position Xt.The acquired laser sensor data is matched with the environmental map and the particles are scored according to the obstacle between the detection obstacle and the obstacle in the map,and the highest score is taken as the estimated position of the robot.In order to reduce the positioning coordinate error,the estimated position is filtered and the filtering effect is evaluated by using the European distance of the estimated coordinates and the real coordinates.MATLAB simulation experiments show that the improved Monte Carlo localization algorithm can effectively reduce the positioning calculation time and has no obvious effect on the positioning accuracy.After the improved Monte Carlo algorithm,the first-order low-pass filtering algorithm is added to improve the positioning accuracy and The impact of positioning time is very small.An improved ant colony algorithm is proposed for global path planning.It includes dynamic updating parameter values,selecting the shortest path of the Nm generation,setting the gate lag value and the "back-off" strategy.The simulation results show that the improved ant colony algorithm can effectively reduce the iteration times and save time.The combination of fuzzy control and DWA(Dynamic Window Approach)algorithm is used to dynamically adjust the weights of DWA sub functions for local path planning.The simulation results show that the improved DWA algorithm can improve the adaptability to the environment and help to find the shortest path quickly.The Robot Operating System(ROS)is used to simulate the navigation robot in Gazebo and RVIZ.The feasibility of improved Monte Carlo location,improved ant colony algorithm and improved DWA algorithm are verified in the navigation simulation.Finally,a mobile robot experiment platform is set up,and the robot control system is combined with the ROS navigation system.The navigation experiment is Completed by constructing the map of the experimental environment.The navigation experiment verifies the correctness of the improved algorithm.