A Universal Positioning System based on Multi-sensor Fusion for Small Mobile Robots

Abstract

For small mobile robots, positioning technology is a prerequisite for ensuring the motion operation. This paper aims to develop a universal positioning system based on multi-sensor fusion, which consists of two parts: hardware and software modules. For the hardware module, two magnetic encoders on a pair of orthogonal wheels and an inertial measurement unit (IMU for short) are used to collect data. For the software module, the Kalman filter is adopted to process data flow to minimize errors and we propose a new coordinate transformation algorithm to adapt to our device. Through the combination of the two modules, we achieve the Orthogonal Wheels System (OWS for short), which can be applied on most kinds of small mobile robots to effectively obtain a relatively accurate position.