Design And Implementation Of Indoor Autonomous Mobile Robot System Based On ROS

In recent years,with the continuous development of technologies such as artificial intelligence,sensors and chip manufacturing,the research of intelligent autonomous mobile robots has also made great progress.A variety of autonomous mobile robots for different purposes are gradually entering human life and work.In a variety of complex applications,how to make the robot to accurately recognise the surrounding,perform real-time positioning and mapping and rapid autonomous navigation have become the research focuses of autonomous mobile robots.This paper focuses on indoor autonomous mobile robots research.The research consists of three areas: robot hardware,control software and application algorithms.Each area is designed and optimized separately to build a self-navigation system that can be used in indoors.The detailed approaches are outlined below:First,according to the actual functional requirements of the robot,a robot chassis based on two-wheel differential drive was selected.The system hardware was divided into four parts:power supply,sensor,control,and host computer system.Each part is optimized.The control part adopts an integrated design,which integrates various functional sub-circuits into a PCB board through an optimized layout approach which increases the reliability and stability of the system;the power supply part is also optimized,and the DC-DC step-down chopper circuit and LDO circuit are used.Power is supplied to the system in three ways,which realizes the power isolation between loads of different attributes and effectively reduces the interference between the systems.This method further reduces the data fluctuation of the sensor measurement,and improves the measurement accuracy.Secondly,the robot's control software is optimized.RT-thread is used as the operating system of the control board.The software is designed in a multi-task mode and task calls are performed in a multi-thread mode,which improves the corresponding speed of the robot control system.Double-circuit closed-loop incremental PID is used to adjust the speed of the motor.The feedback link uses the MCU's CCP hardware to effectively filter the edge pulse glitch voltage.The encoder performs AB phase pulse double-edge sampling to make the feedback data more accurate.Finally,the SLAM algorithm based on laser radar is studied by selecting a particle filter as the main algorithm for robot positioning and mapping.To overcome the problem of particle diversity decline under frequent resampling in the RBPF-SLAM algorithm,the Thompson sampling algorithm is used to compare the traditional RBPF algorithm.The re-sampling process is optimized.And the experimental method has verified the feasibility of the optimization algorithm.A self navigation algorithm based on the known map is being studied.An A* global path planning algorithm and the DWA-based local path planning algorithm are selected as the robot's self navigation algorithm.Through the correct setting of the target point,the robot can plot a smooth and collision-free global path on the map.By performing the real-time obstacle avoidance,the robot has successfully reached the target point.Hence verified the practicability and reliability of the robot system.