| Railway transportation equipment is an important productive asset of railway.Its healthy and stable operation is the important foundation of railway transportation production activities and the guarantee of normal activities of railway transportation organization.Railway power supply monitoring system will collect the thread current,voltage,active power,reactive power,power factor,power extreme value and other technical state data of transportation equipment as an important basis for efficient control and decision-making of railway transportation equipment.In the face of increasing technical state data,the existing methods of monitoring information storage mainly rely on RAID(Red Ex Array of Independent Disks),which uses traditional relational database for management.There are some problems such as poor storage capacity,fixed structure and low reading and writing efficiency,which can not meet the demand of sharing data center business among various application systems.If unable to instant access to the time sequence of quasi real-time massive amounts of data in a quick search of railway power supply tower breaker information,the information of comprehensive analysis with health and safety file information,the monitoring information interaction may cause difficulties,communication congestion,affect the railway power supply production operational efficiency,unfavorable to shorten the troubleshooting time and reduce the rate of the train.Because the traditional relational data management technology based on row storage has the bottleneck of database expansion,the academic research turns to the data processing technology based on column storage.Column data storage has a higher compression ratio than row storage,and more IO operations can be reduced by retrieving specified column data.These advantages are more attractive for storing large amounts of data,and the advantages of using column storage are better,as well as better cost performance.HBase column type database,for example the gathered group processing technology has important applications in some key areas,this article in the railway power supply scheduling monitoring system of mass monitoring information efficiently deal with the hot issues,key research based on HBase column of the compression of the gathered group and optimize the processing technology,and monitoring information were studied based on column storage ratio of compression.On the basis of theoretical research,this paper takes engineering data as an example.Firstly,in order to solve the problems of low transmission throughput and long transfer time between column databases,a memory flattening compression model of distributed data cluster is designed and implemented.Success and then build on HBase gathered group of power scheduling monitoring information processing platform,and will be flat compression model integrated into HBase cluster to achieve Shared memory folding compression,considering the memory fold after optimization performance,taking a train engineering data of SCADA system for example,data loading time,data read time and memory usage rate test multiple sets of data,such as the change of the access to monitoring information of the proposed method is validated through optimization effect.The large ratio compression storage model of monitoring information neural network based on column storage mode is studied and designed,and its compression effect is verified by engineering example data.The research results provide a new solution for improving the reading and writing ability of railway power supply monitoring big data cluster and shortening the processing delay and high ratio compressed storage. |
PDF Full Text Request