Research On Navigation And Positioning Of Robotic Cart Based On Multi-sensor

With the development of modern society,robots are beginning to be used in more and more fields.With the increasing demand of people and the development of current science and technology,indoor mobile robotic cart has become a popular research direction.As an important part of the function of the robotic cart,the navigation and positioning of the robotic cart is getting higher and higher.Especially when the robotic cart is used in an unknown environment,it requires good map construction and path planning functions.The traditional autonomous robotic cart navigation mainly based on single lidar sensors cannot adapt to different characteristics in various environments,resulting in large errors.At the same time,to improve the navigation and positioning efficiency of the robot car,it is necessary to further optimize the existing path navigation algorithm to meet the timeliness and efficiency of its motion process.In order to make up for the shortcomings of the single laser sensor in the autonomous navigation and positioning process of the robotic cart,this paper designs a multi-sensorbased autonomous navigation and positioning system for the robotic cart.The sensors mainly include laser radar,binocular vision camera with depth information and inertial navigation sensor.When the robotic cart is moving in an unknown environment,after multiple sensors process the raw data collected by the sensors,the optimized data is obtained by the Kalman filter-based data fusion method and passed to the robotic cart control system.The control system uses the ROS(ROS Robot Operation System)software platform performs data analysis and map construction,and then builds an environment map of the surrounding environment under an unknown environment.Through the derivation and comparative analysis of relevant theories in the way of building maps,the Cartographer map construction method based on SLAM(Simultaneous Localization and Mapping)is selected.After constructing an environment map in an unknown environment,another important direction for autonomous robotic cart navigation is path planning.Through theoretical comparison analysis and simulation experiment comparison,this subject optimizes the A star algorithm as a global path planning algorithm.Considering the changes in the surrounding environment of the robotic cart running in real time,this subject also chooses the DWA(Dynamic Window Approach)as the local path planning algorithm during the operation of the robotic cart,the efficiency of the path planning process is effectively improved by combining the optimized A star algorithm with the dynamic window planning algorithm.Finally,in order to verify the advantages of autonomous navigation and positioning based on the multi-sensor robotic cart,this paper first builds the system of the experimental car.And then with the environment of barrier-free laboratory and obstructed laboratory,this paper carries out autonomous navigation and positioning experiments on the robotic cart under single laser conditions and multi-sensor conditions separately.Using the map construction effect and the position deviation and time consumption of the robotic cart's docking point and the target point as the evaluation criteria after path planning effect,it is finally proved that the multi-sensor-based robotic cart's autonomous navigation and positioning effect is better than the effect under single laser conditions.