Design And Motion Control Of The Wheeled Mobile Robot Based On Raspberry Pi

Robotics research has been a hot topic all the time.With the continuous improvement of precision of the sensor,computer technology,data processing and level of manufacturing industry,the robotics technology has been developed quickly.As an outstanding representative,the vacuum robot has entered our daily life.Intelligent characteristics and independent movement are the keys to ensure the practical application of intelligent mobile robot.Based on Raspberry Pi,ROS operation system,software and hardware and AI algorithm,the design and implementation of mobile robot and navigation technology are studied in this paper,mainly covers:(1)The ROS operation system is introduced as the software platform,and the overall design of mobile robot system is realized based on ROS.The designed robot includes four wheel chassis driven by two wheels,robot control system constituted by Raspberry Pi,ROS and host computer,robotic perceptual system with laser radar,depth camera,ultrasonic and infrared sensor.(2)The particle filter algorithm is used to locate the robot,and an improved MCL location algorithm is proposed.With two aspects in reducing the location calculation and solving the problem for robot being hijacked,the algorithm introduces a new judgement mechanism to realize the adaption and reduce computation,then the random particle method is added to locate the robot accurately.(3)Combining the global path planning with the local path planning,the optimal path planning of the robot in the running environment is realized.The A* algorithm is selected as the method of the global path planning.In addition,the situations that robot may encounter with during running is fully considered to realize the robot's local path planning and real-time obstacle avoidance with fuzzy control theory.In the indoor environment,a variety of situations which the robot may encounter have been simulated and verified by experiments,and the feasibility of the algorithm is tested.(4)The designed robot control system and the proposed navigation algorithm are tested.In different indoor environments,the performance of robot is tested,which includes mobile error,accuracy of map building,location and navigation error,etc.The results of experiments are carefully analyzed to improve the parts where the large errors of robot exists.