Forecast Method Study on Track Longitudinal Level Irregularity Based on Most Possible Similarity Model
Track irregularity is objective existence, it cant be eliminated. But, we can formulate reasonable repair and maintenance plans to increase the efficiency of track maintenance and ensure safety, if the prediction of track irregularity is effective. In this paper, on the basis of analyzing on the irregularity forecast method and the "memory" of ballast track, a method based on most possible similarity model is introduced. This method is going to find irregularity in history most close to the current, and use its development to predict the current development. The longitudinal level irregularity of Hefei-Wuhan line was predicted, and results indicate that prediction graphics were consistent with the actual graphics in trend. So, using this model to predict development trend of track longitudinal irregularity is feasible.