Multisensor Integration Methods in the Development of a Fault-Tolerant Train Navigation System

Abstract

Onboard train positioning (navigation) plays a vital and safety critical role in advanced Automatic Train Control (ATC) and Automatic Train Protection (ATP) systems. Such onboard systems are also essential for moving block signalling and control systems for railways. The application of multi-sensor fusion algorithms to the vehicle navigation field has made it possible to create inexpensive and accurate positioning systems, which will satisfy the railways' requirements. The state estimation methods involved in Kalman filtering have proved to be some of the most effective techniques in multi-sensor data fusion. A multi-sensor navigation system is introduced in this paper to address the shortcomings of the existing train positioning systems. The proposed system utilizes the Global Positioning System (GPS), Doppler radar, gyroscopes, tachometers, digital maps and balises. In order to provide fault detection and isolation capabilities, a hierarchical structure is proposed for the multi-sensor integration system in which different combinations of navigation systems would function. Several data integration nodes, including DR/GPS, DR/Balise, and DR/GPS/Balise, are studied in more detail and their performances are evaluated.