Research on prediction method of life cycle of track quality based on grey system identification

The grey nonlinear model put forward by the author (see the Journal of the China Railway Society, 2010, 2) is viewed as a time function controlled by a series of parameters including trend coefficients and random coefficients. The coefficients are identification parameters for system analysis, which represent the characteristics that the system is varying with time. The prediction model is established under the conditions of knowing tamp operating efficiency of large machine or initial quality of railway track. The relevance among track quality systems in different maintenance periods is assumed, and then TQI characteristics in the maintenance period required predicting can be obtained by data mining on identification parameters of TQI in the known maintenance periods. Through the calculation example we can see the prediction model proposed in this paper can predict the development of track quality in different maintenance periods accurately and also provides a new idea for the research on life cycle of track quality.