Multi-ultrasonic sensor fusion based on Dezert-Smarandache theory

It is well known that ultrasonic sensor is widely used in mobile robot to get range information with low cost, and the main bottle problem of using ultrasonic sensors to build an occupancy grid map is how to deal with the uncertainty in ultrasonic sensor data. In this paper, a new kind of Multi-ultrasonic sensor fusion method is presented. The ultrasonic sensor model is built with fuzzy set theory to describe the uncertainty contained in sonar data. Environment information obtained from different measures is processed with Dezert-Smarrandache (DSm) theory which is a new methodology for solving fusion problems. To get a high quality occupancy map, a paradoxical factor (PF) is introduced to help to deal with multi-reflection of ultrasonic sensor. Through simulation and experiment results, it is proved that the new algorithm behaves a good performance during mobile robot map building.