Adhesion Characteristic Slope Estimation for Wheel Slip Control Purpose Based on UKF

Abstract

Slip controllers are crucial for railway vehicle operation when a high tractive effort is required. The slip controllers have been developed for decades, and their developments continue. The developed slip controllers are based on many more or less effective principles. One of the perspective slip controllers determines a phase shift between an injected excitation signal to the required torque of the electric drive and the signal representation in a measured wheel angular velocity. The slip controllers that are based on the principle can effectively work with any adhesion conditions. However, the slip controller principle has some drawbacks like high filtration requirements, possible failure due to mechanical shock in the wheel-rail contact and long response time. The paper proposes a novel phase shift detection method for the slip controller based on an unscented Kalman filter that can cope with the known disadvantages of the original slip controller. When the proper output of the UKF is used, the slip controller becomes faster, more reliable and can work when the original slip controller fails. The proposed slip controller was verified on the measured data in an open-loop calculation and the mathematical model in the closed-loop simulation.