Research And Implementation Of Distribution Network Monitoring Method Based On Big Data

With the development of the country’s power system,the requirements for the power system are getting higher and higher,and the distribution network is a key part of the power system.In the distribution network,however,there are problems that cannot be ignored.Smart meters with full monitoring point coverage make the amount of voltage qualification rate monitoring data expand dramatically,in addition to heavy loads and overloads in distribution transformers,which seriously endanger the safety of electricity consumption.Therefore,in order to solve the problems of voltage compliance rate monitoring and distribution transformer heavy overload monitoring within the distribution network,this thesis proposes a corresponding solution and verifies its effectiveness through experiments.The main work and contributions of this thesis are as follows.(1)To address the problems of large data volume and variable scenarios in the distribution network pass rate monitoring task,this thesis proposes a parallel aggregation calculation method based on Flink.The method consists of five modules: data acquisition,data pre-processing,feature fusion and conversion calculation,parallel aggregation calculation,and data merging.The method combines the practicality of the power industry and improves the computational efficiency of monitoring and analysis through hot data caching and parallel aggregation computation,and can be applied to a variety of monitoring scenarios.Experiments show that the task time is reduced by 28% compared with the original method of a power grid company.(2)In response to the problems of various causes of heavy overloads in distribution transformer monitoring tasks,the difficulty of analysis and the lack of monitoring timeliness,this thesis proposes a real-time heavy overload monitoring method based on Flink.The method consists of five modules: data acquisition,real-time power data reading,load factor calculation,state calculation and state output.The method not only solves the heavy overload monitoring analysis problem,but also has more timeliness compared with the original monitoring method of a power grid company.Experiments show that the method meets the requirements of real-time heavy overload monitoring scenarios in terms of overall throughput and computation latency,which is much better than the original monitoring method.(3)Finally,this thesis carries out a data visualisation display based on the Vue front-end framework to summarise the daily pass rate and pass rate trend in the voltage pass rate monitoring task as well as the real-time abnormal results in the heavy overload monitoring task,which is convenient for analysing the problem from a macro perspective.Similarly the work of this thesis is more complete monitoring and analysis results are given to the power maintenance personnel to facilitate their follow-up processing of the problem,ultimately achieving the purpose of being able to accurately identify and analyse problems within the distribution network.