基于轨检车历史数据的轨道不平顺预测

Prediction of Track Irregularity Based on Historical Data of the Track Inspection Car

由轨检车检查得到的大量轨道几何尺寸的历史数据,可以挖掘出轨道不平顺发展背后的潜在规律。从已有的历史数据的角度出发,以数据驱动的方式为出发点,得出轨道不平顺的发展趋势,可以做到对轨道质量指数的合理预测;若单纯从某一理论方法出发进行研究,实际观测数据不一定都能符合其结果。基于已有的历史数据出发,寻找最为匹配的规律,为轨道恶化趋势的预测提供新思路。研究结论如下:(1)提供了对轨道质量预测的新思路,预测结果的置信度可达到约92%;(2)基于已有的原始数据,匹配得出的分布函数中的系数根据实际情况的变化而变化,可包容被测轨道的不同个性;(3)轨道质量指数的增长与时间的关系是一种理论上循环的非线性关系。

A large number of historical data of track geometry by the rail inspection car can excavate the potential law behind the development of track irregularity.With the existing historical data,the method of data driven is the starting point,the development trend of track irregularity is obtained,reasonable prediction of track quality index can be achieved;If it is studied simply from a theoretical approach,the actual observed data do not always correspond to the results,based on the existing historical data,the most matching law is found,it provides a new idea for predicting the deterioration trend of track.Research conclusion are as following.①A new idea for prediction of track quality is provided,and the confidence level of prediction result can reach about 92%.②Based on the existing raw data,the coefficients in the matched distribution function are changed according to the actual situation,and the different individuality of the measured track can be contained.③The relation between the increase of TQI value and time is a nonlinear relation in theory.