Research On High-speed Railway Disaster And Intrusion Monitoring Technology

Railway is the main artery of national economy,which has effectively promoted China's economic and social development.In particular,the opening of more and more high-speed railways has greatly facilitated people's travel and changed people's lifestyle to some extent.With the expansion of high-speed railway network and the operation of major high-speed railway lines at the speed of 350km/h,the safety of train operating environment has become a major problem in the safe operation of high-speed railway.The main manifestations are that natural disasters occur frequently along the lines,hidden dangers in the surrounding environment have always existed,and man-made sudden disturbances have occurred from time to time.Once a disaster occurs,it will seriously interfere with the order of train operation,threaten the safety of passengers' lives and property and greatly affect the social stability.In China,the natural disaster and intrusion monitoring system and earthquake earlywarning and monitoring system have been constructed simultaneously with the construction of high-speed railway,while perimeter intrusion alarm system,line obstacle automatic monitoring system and slope monitoring and early warning system are under development.Real-time monitoring,alarm,and early warning for natural disasters such as wind,rain,snow,earthquakes,and foreign body intrusion,personnel intrusion,landslides,and debris flows,etc.,which have a great impact on train operation,ensure the safety of train operation.However,the above system has the following problems: lack of unified planning and independent settings,resource duplication,inconvenient to use and maintain.The monitoring data is not fully analyzed and utilized.The ability to identify abnormal wind and rain data is insufficient.The layout of some monitoring points is unreasonable.The wind warning capability is insufficient.The handling of the wind alarm is delayed.The all-time small target intrusion monitoring technology research is insufficient and so on.To solve the above problems,in this dissertation,the architecture design of the highspeed rail disaster and intrusion limit monitoring system based on cloud native is carried out,and the identification technology of high-speed rail wind and rain anomaly monitoring data and the optimization method of wind monitoring point optimization layout are proposed.Rapid disposal technology,a railway perimeter intrusion monitoring technology based on multi-sensor fusion and a railway intrusion detection algorithm for small targets are proposed.This paper mainly conducts the following four researches:(1)In terms of top-level design,the existing problems such as system independence,repeated construction and lack of sharing are analyzed and the overall architecture,logical architecture,technical architecture,data architecture and system functions of the cloudnative high-speed railway disaster and intrusion monitoring system are proposed.It realizes the resource integration and data sharing of various systems,and can flexibly expand other monitoring business systems.(2)In terms of data collection and processing,focusing on the inaccurate identification of abnormal wind and rain data in the application of disaster monitoring systems,the identification technology of abnormal data in wind monitoring based on correlation and difference test and the identification of abnormal data in rainfall monitoring based on LSTM are carried out.Technical research improves the accuracy of wind and rain alarms.In the layout of wind monitoring points,the optimization method of new and existing lines based on the combination of historical meteorological data and railway terrain and lines has been carried out to make the layout plan of monitoring points more reasonable.(3)In the aspect of high wind early warning and disposal,in view of the nonlinear and non-stationary characteristics of wind speed,a wind speed prediction algorithm along the high-speed railway lines based on CWS(CEEMDAN-WOA-SVR)technology is proposed.Using this method can reduce the influence of nonlinear and non-stationary wind speed time series on the prediction results.The operation rules and rapid disposal technology of gale disasters were studied to ensure the safe operation of high-speed railway.(4)In terms of intrusion limit monitoring,because the current single technology cannot meet the requirements of ‘zero false negatives and low false positives',different scenarios and technologies are analyzed,and the multi-sensor fusion limited monitoring program based on ‘one scene,one case' is proposed.Aiming at the problem of small target recognition,a detection algorithm using an improved CMR network is proposed,which improves the practicability of intrusion limit monitoring.