Online Estimation Algorithm of Adhesive Force for Train Wheeltrack

Abstract:

Novel online estimation algorithms of train adhesive force were proposed for antiskid control. Since it is difficult to measure the adhesive force, it is of significance to make full use of adhesive force in antiskid control. In this paper, a wheel set dynamic model is established first. Then, using the Calman filter, extended state observer and so on, five online estimation algorithms for train adhesive force were designed, where the axle speed and the equivalent clamping force were the input. Furthermore, with the simulink software platform, signal noise contamination and transmission delay were set, and two conditions of constant adhesive force and variable adhesive force were simulated. Simulation results reveal that the five algorithms could be used to estimate the adhesive force, but when the response time and the maximum error of adhesive force estimation are taken into account, the nonlinear expansion state observation algorithm is the best algorithm for adhesive force estimation. Finally, the accuracy of the estimation algorithm is further validated by using measured data.

Key words:adhesive force; state observation algorithm; wheelrail relationship; antiskid control