基于劣化趋势模型的轨道病害识别方法研究

Research on Track Disease Identification Method Based on Deterioration Trend Model

摘要: 道床病害是引起有砟轨道线路轨道平顺状态恶化的重要原因。针对这一问题，本文基于有砟铁路大量轨检车历史检测数据构成的非等距时间序列，提出一种BP时间序列分段法，并在此基础上建立轨道高低不平顺标准差劣化率模型，通过轨道高低不平顺劣化速率间接筛选道床劣化严重区段，并通过现场复核对进行了验证。结果表明：轨道高低不平顺标准差随维修呈周期性劣化趋势，同一区段捣固等维修作业前后劣化率变化不大，且在同一劣化周期内，轨道高低不平顺近似线性趋势劣化；但当道床发生病害时，轨道高低不平顺标准差劣化率急剧增加，并可能与相邻区段劣化率有所差异。本文通过对轨检车动态检测轨道不平顺数据进行挖掘可以快速有效识别道床病害区段，对现场道床养护维修具有指导意义。

Abstract: The ballast-bed disease is an important cause of track smoothness deterioration of ballasted track lines. To solve this problem, this paper proposes a BP time series segmentation method based on a non-equal interval time series constructed from a large amount of historical detection data from railway inspection vehicles. On this basis, a standard deviation degradation rate model for track unevenness is established. The severely degraded section of the ballast-bed is indirectly screened by the deterioration rate of the track, and verified by on-site review. The results show that the standard deviation of track irregularity tends to deteriorate periodically with maintenance, and the deterioration rate does not change much before and after maintenance operations such as tamping in the same section, and the track irregularity tends to deteriorate approximately linearly within the same deterioration period. However, when the ballast-bed disease occurs, the deterioration rate of the standard deviation of track irregularity increases sharply and may be different from that of adjacent sections. By mining the data of the track irregularity detected by track inspection car, the paper can quickly and effectively identify the diseased section of the ballast-bed, which has guiding significance for the maintenance and maintenance of the ballast-bed on site.