Robust Real-time Obstacle Avoidance of Wheeled Mobile Robot based on Multi-Sensor Data Fusion

Abstract

Wheeled mobile robot has the advantages of strong execution ability, stable operation and fast response speed. It is widely utilized in smart autonomous robots, automated operation platforms, urban rail transit and some other related fields. Currently, the most of researches on wheeled mobile robot autonomous motion are based on single-sensor data to detect surrounding objects, which are susceptible to the environmental interference. This paper adopts the centralized multi-sensor fusion algorithm to achieve obstacle avoidance based on Kalman filter. Through multi-objective optimization, all the sub-behaviors among movement process are weighted, which improves the real-time and robustness of the system. Finally, the effectiveness of the algorithm is verified through prototype testing.