A multi-robot positioning system using a multi-code ultrasonic sensor network and a Kalman filter

In multi-robot systems, each robot needs to have the position and pose information of itself and that of the other cooperative robots. This paper presents a synchronous distributed positioning system that uses a multi-code ultrasonic sensor network and a compensation algorithm using a Kalman filter. The bearings of robots are computed by using their position changes, and then compensated for by using the Kalman filter. The ZigBee sensor network protocol is used for communication among the robots and for the synchronization of the ultrasonic transmission timing. The experimental results show that our system positions multiple robots synchronously without any configured infrastructures. The results have a better accuracy and less accumulative error than those found in positioning systems without compensation.