Researsh On Real-time Processing Of Railway Power Supply Monitoring Based On Distributed Kafka Queue And Stream Computing Cluster

The railway power supply system is an important equipment for railway transportation and is responsible for the power supply for electric locomotives,on-street stations,signal lights,and blocking devices.The safety of power supply will directly affect the safety of railway transportation.To ensure the safe power supply of railway systems,The railway department dispatches and monitors the railway power supply system through the railway power supply monitoring system.The railway traction power supply system is a typical dynamic electricity network,and the locomotive load is a high-power impact load,which has the characteristics of large fluctuations,rapid changes,etc.At the same time,the high-speed train has a fast operation speed and a large open-line density,making the railway power grid Voltage,current,and other operating parameters change frequently.The monitoring terminal uploads the collected live real-time monitoring information to the dispatch center.As the number of on-site measurement points increases and the acquisition frequency increases,the amount of power supply scheduling information collected is also increasing.More,long-term operation will generate huge amounts of information.The traditional data processing mode is likely to cause a large processing delay when processing massive monitoring information,which not only affects the real-time performance of the railway power supply dispatching system,but even results in the loss of key monitoring information in severe cases,resulting in the failure to process the fault information in time,thus threatening.Power supply to the railway system.Therefore,how to quickly process massive amounts of real-time monitoring data is a key issue that needs urgent solution.This paper aims at real-time processing of massive monitoring information in railway power supply monitoring system,and combines with the processing requirements of massive monitoring information in practical engineering applications to set up a real-time processing system for railway dispatch monitoring flow calculation.The system uses Storm flow calculation framework as real-time monitoring data.The processing module uses the distributed message system Kafka as the data cache module and takes the data of the 10 kV electric remote monitoring system of the railway car segment as the test object.The real-time processing system for convection calculation performs a series of topological example tests of the monitoring data,and the data of the system.The delay performance and data throughput performance are tested.The test results show that the real-time processing system for flow calculation can obtain 100 ms-level processing delay,and has good data throughput performance.It is a new solution for the rapid processing of mass data in railway powersupply monitoring.At the same time,in order to improve the overall computing capability of the system,the real-time processing system for flow calculation is optimized from the aspects of data caching module,parallelism parameters of topological components,and task scheduling methods.Through comparison experiments before and after optimization,it is verified that the optimization scheme improves the flow calculation and processing system.and the load of the cluster work nodes is more balanced when running the task.