Research on Omnidirectional Indoor Mobile Robot System Based on Multi-sensor Fusion

Abstract

Indoor mobile robot faces the problems such as complex situations, multiple barriers, narrow spaces and low navigation accuracy. In this paper, omnidirectional indoor mobile robot system based on multi-sensor fusion is established which consists of mobile robot body, upper order control system and lower order control system. The mobile robot body is a platform adopting mecanum wheels as actuating mechanism. The upper order control system consists of upper computer controllers and robot operating system (ROS) above the controllers. The lower order control system consists of chassis controller using STM32F103 chip as the core part, DC motors and encoders. The navigation and localization function are implemented by data of sensor such as laser radar, UWB, odometer and IMU. The upper order control system first fuses and programs the navigation and localization data from multiple sensors, then sends instructions to the lower order control system; the lower order mobile control system receives these control instructions and solves the kinematic model, then control the robot body using PID algorithm. Localization and navigation test environment of the omnidirectional indoor mobile robot is established. Experimental results show that localization and navigation of omnidirectional indoor mobile robot system based on multi-sensor fusion has the advantages of omnidirectional indoor mobile ability, strong barrier avoiding ability, and high localization and navigation accuracy.