Research On Key Technologies Of High-speed Railway Emergency Management Driven By Big Data Analysis

With the continuous expansion of high-speed railway lines in China,high-speed railway takes on more and more important passenger transport tasks.High-speed railway has the characteristics of fast running speed and small driving interval.The occurrence of emergencies will not only affect the running organization of high-speed railway,but also bring great harm to the safety of life and property of passengers.Therefore,how to apply new information technologies such as big data to rapidly realize high-speed railway emergency management and minimize the impact of emergencies on high-speed railway passenger transport is crucial.Around the high-speed railway emergency management,this paper puts forward the cloud service oriented high-speed railway emergency management platform and the overall framework of the double particle swarm of high-speed railway cloud resource scheduling method,and the key of high-speed railway emergency data integration management,high-speed railway emergency plan optimization and high-speed railway emergency plan are studied,such as digital finally in railway engineering application,for example,has carried on the example is given to illustrate and verify.The main innovation points of this paper are as follows:(1)Based on big data,cloud computing,Internet of things and other new technologies,this paper proposes the overall framework of high-speed railway emergency management platform oriented to cloud service,covering four stages of railway emergency management,prevention,preparation,response and recovery.Resource scheduling problem of emergency services in the cloud model,according to the service dispatch needed,such as time,cost and availability,based on cloud services for all kinds of internal and external resources for unified management and coordination,to provide the corresponding with the characteristics of high speed railway emergency disposal cloud service combination scheduling,implementation way to emergency linkage.(2)Based on the knowledge map of multi-source heterogeneous foreign high-speed railway emergency data organization scheme,this paper applyies the knowledge mapping technology to build the knowledge map,high-speed railway emergency big data based on Hadoop framework of high speed railway emergency big data platform overall architecture,using the RSA encryption algorithm for high speed railway emergency sensitive data is encrypted,and put forward the fusion method of AHP and DMM emergency big data maturity of high-speed railway are analyzed.(3)Analyzed the high-speed railway emergency plan system and high-speed railway emergency response hierarchies,high-speed railway emergency plan optimized data algorithms are given,and according to the actual data conducted based on the test of the high-speed catenary inspection data analysis plan optimization,based on the mean shift clustering of train fault and the interruption of driving time correlation analysis,and based on the decision tree analysis of electricity failure associated with the interruption of driving time,for high-speed railway emergency plan optimization,and incident response level segmentation,etc.To provide decision support.(4)The overall framework of digital management of high-speed railway emergency plan integrating ontology and deep learning is proposed,which realizes the annotation and word vector generation of high-speed railway emergency information,and realizes the extraction of key fields such as time,place and emergency name through bi-lstm +CRF deep learning model.Using ontology,high-speed railway emergency plan is constructed digitally,and based on object tree method and semantic matching and searching of ontology,high-speed railway emergency disposal process is generated automatically.Finally,this paper takes the high-speed railway emergency management of railway bureau as an example and applies Java SSH architecture to build the high-speed railway emergency management platform of railway bureau and realize engineering application.