Research And Design Of Autonomous Navigation System For Multi-Sensor Fusion Disinfection Robot

Due to the impact of the COVID-19,hospitals and other public places require repeated disinfection.Manual disinfection greatly increases the risk of infection and virus transmission.Using robots to replace manual disinfection is a relatively safe and effective approach.Therefore,autonomous navigation disinfection robots have attracted much attention.Currently,autonomous navigation disinfection robots have some shortcomings,such as single sensor mapping,incomplete navigation information,and insufficient accuracy.Therefore,researching autonomous navigation systems for disinfection robots using multisensor data fusion has significant research value to improve the overall performance of disinfection robots.This paper studies an autonomous navigation system for disinfection robots based on multi-sensor data fusion.The system effectively builds indoor environmental maps and achieves autonomous navigation through data fusion from twodimensional laser radar and depth camera.The main contributions of this paper are as follows:(1)The motion mode and function of the autonomous navigation system for disinfection robots are analyzed.Based on the analysis,the mobile chassis of the disinfection robot autonomous navigation system is constructed,and the hardware and software framework of the system is designed.The kinematics,coordinate system,odometer,laser radar,and depth camera models in the navigation system of the disinfection robot are established.(2)Addressing the limitations of single-sensor mapping in complex working environments for disinfection robots,we propose a multi-sensor fusion-based mapping approach.This approach converts depth camera data into pseudo lidar data,fusing it with lidar data at the data level to obtain environmental information data.We fuse attitude sensor data and wheel odometry data through extended Kalman filtering to obtain robot pose information data.The environmental information data and robot pose information data are used as inputs for Synchronous Localization and Mapping(SLAM)based on the Cartographer algorithm.We conducted simulation experiments in the Gazebo environment to verify the feasibility of the proposed mapping approach.(3)The positioning and navigation algorithm of the disinfection robot is studied,and the navigation scheme of the disinfection robot is built.Adaptive Monte Carlo Localization(AMCL)algorithm was used for robot localization,A* algorithm was used for global path planning,and Time Elastic Band(TEB)algorithm was used for local path planning.The obstacle avoidance function is completed by subscribing the data fused by the lidar and depth camera.Simulation experiments are carried out in the map constructed by the multi-sensor fusion mapping scheme,and the navigation effect of the scheme is verified.(4)An indoor experimental scene is set up to test the mapping and navigation function of the designed autonomous navigation system of the disinfection robot.Through the comparison experiment of single sensor mapping and multi-sensor fusion mapping scheme,the experimental results show that the map constructed by the multi-sensor fusion mapping scheme proposed in this paper is more complete and accurate,which meets the needs of disinfection robot indoor environment scanning.The experimental results show that the system can complete the autonomous navigation task,carry out global and local path planning,avoid unknown obstacles in the environment in real time,and safely reach the navigation target point.