Modeling of Train Braking Based on Environment and Online Identification of Time Varying Parameters

Abstract

In this paper, the mechanism of pure air emergency braking for high speed train is analyzed. Considering the influence of the actual train running environment on the braking performance, an emergency braking model based on the environment is established. The expectation maximization identification of braking model for high speed train based on sliding window is proposed, and the unobserved time varying adhesion coefficient is identified. Firstly, the position and size of the sliding window are determined. Then the adhesion coefficient is identified by expectation maximization based on sliding window. Finally, combined with gradient optimization, the optimal identification of adhesion coefficient is obtained. The simulation results show that the online identification proposed in this paper can be used to identify the adhesion coefficient quickly and accurately. Under uniform noise, the identification error and relative error of adhesion coefficient are ±0.0015 and 1.8705% respectively. The relative error and root mean square error of braking speed are 0.4038% and 0.1018 respectively. It is satisfied with the actual needs of the braking system, and the accuracy of the model and effectiveness of the online identification method can be verified.