Innovative Management of Wheel-Rail Adhesion Conditions in Localization Algorithms for the Automatic Train Protection

Abstract

In the modern railway network, Automatic Train Protection and Control (ATP-ATC) systems are fundamental to increase the infrastructure capacity, maintaining a proper level of operation safety. Odometry is a relevant on-board module, since it estimates the speed and the travelled distance of a railway vehicle along the track. Its reliability affects the definition of allowed speed profiles, in order to prevent collisions from a driver's failure to observe a signal. Typically the dead reckoning relies on encoders providing an accurate estimation of the train speed only when good adhesion conditions between wheel and rail occur. In presence of wheel sliding the estimation error may become very high.
In this paper the management of adhesion conditions is enhanced using an Inertial Measurement Unit (IMU) which improves the adhesion condition detection and the speed and travelled distance estimation.
The testing of the proposed algorithm is performed through a testing simulator, set up by the MDM Lab and used to speed up the algorithm tuning, capable of replicating in a realistic way the motion dynamic effects of a railway vehicle on inertial sensors. The Hardware-In-The-Loop test rig is composed of a Matlab-Simulink^TM three-dimensional multibody model of a railway vehicle, a commercial anthropomorphic manipulator with spherical wrist and an IMU designed by ECM Spa (Pistoia, Italy). The experimental results are compared to the performance requirements fixed by the European Rail Traffic Management System (ERTMS).