摘要
为了保证列车在轨道上安全运行,铁路部门采用动态检测和静态检测两种方式对铁路轨道进行检查。笔者针对现有铁路轨道检测方式,提出根据检测数据评定轨道质量状态的方法。该法采用数据仓库技术存储检测数据,给出检测数据存储的双重粒度模式和数据分割模式。从时间和空间两个方面对检测数据进行分析,采用轨道局部状态、整体状态、超限点重复率和重点地段分析的方法。将轨道数据分为时间维、地理维、类别维3个维度,运用星型模式对轨道检测数据进行数据挖掘,发现轨道变化规律。将上述方法应用到南疆工务管理信息系统中,取得了较好的实用效果。
Railway departments use dynamical and static detection methods to detect the track condition so as to ensure trains run on it safely. Based on current detecting methods, a new method is put forward to assess the quality of railway track according to detected data, which uses data warehouse to store the detected data with double-granularity mode and data partition mode. In this method, the detected data are analyzed respectively from the angle of time and space. Methods for evaluating the local and global track conditions and for calculating the repeat ratio of ultra-limit spots and key sections are given. The detected data, divided into temporal dimension, spatial dimension and classificatory dimension, are mined by using star mode so as to discover the changing laws of track. These methods are proved to be effective in practice after applied to the Naniiang operational management system of Urumchi Railway Bureau of China.
出处
《中国安全科学学报》
CAS
CSCD
2007年第4期166-171,共6页
China Safety Science Journal
基金
北京交通大学校基金资助(2005SM029)
关键词
检测数据
轨道状态
评定
数据仓库
数据挖掘
detected data
railway track condition
assessment
data warehouse
data mining
