On Using Unscented Kalman Filter Based Multi Sensors Fusion for Train Localization

To guarantee high precision control and assure the reliability and safety of the railway systems operation, highly accurate measurement data from different onboard and wayside sensors are needed. This paper proposes a solution approach to obtain more accurate measurement data which are collected during train movement using an Unscented Kalman Filter (UKF)-based multi sensors data fusion. Several sensors that are used in the proposed approach include indoor positioning systems, radio frequency identification, and rotary encoder. Experimental results that were conducted on a train model show how the proposed method effectively able to provide more accurate measurement data about the train position.