Health Status Prediction of High-Speed Railway Wheel Diameter Size Based on Hidden Markov Model

Abstract

It is beneficial for reducing maintenance costs and improving train safety and reliability to predict and evaluate the condition of railway wheel-set dimensional value. A wheel-set size state prediction model based on the Hidden Markov Model (HMM) algorithm is proposed to predict the state of high-speed railway wheel diameter. The state interval is established according to the change range of the wheel diameter value. And then the corresponding transition probability matrix is used as the input data of the HMM to establish a state prediction model of the wheel size. Finally, the state prediction result is obtained by this method. The experimental results show that aiming at wheel-set data, the accuracy, precision, recall rate and Fl of HMM can reach 0.9, 0.8, 0.8 and 0.8 respectively. Compared with Markov, Graymarkov based on state prediction algorithm and Anfis algorithm based on data prediction, the validity and accuracy of the HMM model in predicting wheel-set size state are verified.