A robust method of fusing ultra-wideband range measurements with odometry for wheeled robot state estimation in indoor environment

摘要

The Global Positioning System (GPS) is a popular satellite-based positioning system which provides world-wide coverage to allow us obtain precise position information in outdoor environment. Today, the number of applications which relay on indoor state estimation is rapidly increasing. Ultra-Wideband (UWB) technology provides a substitution in indoor environment due to its several advantages such as low-cost, high-precision, easy-deployment. However, it can't provides information of orientation. Wheel encoder is a common device that converts wheel's angular motion to a digital signal. With this signal, robot's posture can be easily calculated by odometry in it. This paper describes a method using Particle Filter for fusing UWB range measurements with odometry to acquire a real-time posture estimation of a wheeled robot in indoor environment. We improved the firmware of UWB devices and trilateration algorithm for higher measuring stability when in the presence of obstacles. Performance is experimentally investigated in several scenarios and static positioning accuracy is ±5 cm, dynamic positioning accuracy is ±10 cm.