Prediction Of Metro Wheel Wear And Optimization Of The Wheel Re-profiling Strategy

With the fast development of urban rail transit, it puts forwards a higher request on trains, tracks and related equipment maintenance and life management. As a key component of metro, metro wheel plays a vital role for traffic safety and operation stability. Due to the continuous wear of wheels when moving, maintenance managers have to inspect wheels frequently and re-profile or replace them when necessarily for ensuring traffic safety. The maintenance cost of wheels plays a key role in the maintenance expenditure of rail transport. Therefore, the study of prediction methods of metro wheel wear and optimization of the wheel re-profiling strategy and making reasonable re-profiling strategy, which not only contribute to discovering the safety trouble of wheels promptly, but also is of great significance to extend life of wheel and reduce the maintenance cost.This paper is based on analyzing the prediction of wheel wear and the research status of optimization on wheel re-profiling strategy at home and abroad, and also on analyzing their characteristics. First, with the application of the theory of time series, ARIMA(p,d,q) forecasting model of flange thickness and wheel diameter accumulative wear data is built respectively. The analysis results of example show that the proposed prediction method is effective for the short-term prediction of the metro wheel wear.Second, for the research of optimization of single wheel re-profiling strategy, the wear states of flange thickness and wheel diameter are divided based on statistical analysis of the wheel wear data. And then with the application of the theory of Markov processes, the wear model of flange thickness and wheel diameter is built. Utilizing control limit policy for the wheel re-profiling decision and the Monte Carlo simulation, the wheel re-profiling strategy is optimized. And four kinds of preferred options of the wheel re-profiling strategy are proposed, which are effective for extending life of wheel. Besides, a method of optimization about re-profiling strategy based on a kind of data-driven model is proposed for eight wheels belong to the same carriage. In the case of considering safe range of flange thickness and wheel diameter, the diameter difference among eight wheels and other constraints, a data-driven model of the wheel diame-ter, the flange thickness, and the re-profiling gain is built based on their probability and statistics relations. With the application of the theory of genetic algorithm(GA) and Monte Carlo simulation, optimization of re-profiling strategy is realized, which takes eight wheels belong to the same carriage into consideration. The simulation results show that the preferred re-profiling strategies suggested by this study can significantly increase the life length comparing with the existing re-profiling strategies.Finally, for the deficiencies of the existing wheel data management and record mode, the metro wheel maintenance management software based on Labview is developed according to maintenance management requirements of the metro wheel and maintenance managers’ habits. Its functions include data access, data query, statistical analysis, fault alert, wear prediction and so on, which meet the basic requirements of the maintenance managers.