Research And Implementation Of Robot Synchronous Positioning Mapping And Path Planning In Indoor Environment

With the development of intelligent sensing and control technologies,map construction and path planning for indoor mobile robots have been widely used in indoor navigation,automated warehousing,environmental monitoring,etc.In the study of map construction for robots,commonly used methods include laser SLAM(Simultaneous Localization and Mapping)and visual SLAM.Among the laser SLAM techniques,the more mature algorithms include Gmapping,Hector and Cartographer.In robot path planning research,commonly used algorithms include search-based algorithms such as A* algorithm and Dijkstra algorithm,and sampling-based algorithms such as RRT(Rapidly-exploring Random Trees)and PRM(Probabilistic Road Map).The ROS provides a complete robot operating system with good communication mechanisms and a library of tools to facilitate the development and design of robot systems by robot developers.The map building and path planning capability of indoor mobile robots is one of the key technologies for their intelligence.Currently available navigation methods are still inadequate in achieving dynamic unknown scene building and dynamic obstacle avoidance in the face of unknown dynamic scenes and obstacles.This thesis investigates the problem of map building and path planning for indoor mobile robots based on ROS,taking indoor mobile robots as the research object.The main research elements,are as follows:(1)The ROS robot operating system and its communication mechanism are described,and the overall system of the mobile robot is designed,including three layers: the human-robot interaction system layer,the software-driven system layer and the hardware execution system layer.Then,the hardware selection was carried out,including the main controller,drive motors,IMU(Inertial Measurement Unit),odometer,infrared sensor and LIDAR.Then,the two-wheel differential crawler model is selected as the chassis model,and the kinematics modeling and analysis of the mobile robot are carried out.(2)Commonly used SLAM algorithms are introduced,and the Gmapping algorithm,Hector algorithm and Cartographer algorithm.By improving the resampling strategy,a Gmapping optimization algorithm based on the improved resampling strategy is proposed,which has high fitting accuracy.Combining the characteristics of Gmapping and Cartographer algorithms in mapping,a synchronous positioning and mapping method that combines Gmapping and Cartographer algorithms is proposed,which can improve the efficiency of mapping while ensuring the accuracy of mapping.(3)In the path planning,the thesis divides the path planning into two aspects: global path planning and local dynamic obstacle avoidance,the A* search algorithm is selected for global path planning,and it is improved by optimizing heuristic functions and curve smoothing methods.In terms of local dynamic obstacle avoidance,the Dynamic Window Approach(DWA)is used.This method makes the inflection point of the path trajectory less and smoother,and increases the adaptability of the mobile robot to a certain extent.(4)A virtual simulation environment is built in Gazebo,and the laboratory environment is selected as the physical scene,and the map construction and path planning experiments of indoor mobile robots are carried out.It is verified that the Gmapping optimization algorithm based on the improved resampling strategy and the synchronous positioning and mapping method that integrates Gmapping and Cartographer algorithms have better mapping effects;the A* search algorithm combined with the DWA local dynamic obstacle avoidance method,after heuristic function optimization and curve After smoothing,it can better complete the path planning task in the case of known map and unknown map.