Research On The Prediction Of Emergency Repair Work Order For Distribution Network Based On Big Data Technology

State grid has always adhered to the corporate purpose of "people's electricity industry for the people",and carried forward the core values of customer-centered,professional focus and continuous improvement.In order to comprehensively enhance the sense of gain and happiness of the people,and improve the power supply capacity and reliability of urban and rural areas,power supply must be restored in the first time after the occurrence of power failure.If the order quantity of distribution network emergency repair can be predicted and studied,the emergency repair resources can be optimized in advance,the time of emergency repair can be greatly shortened,and the service capacity of power grid companies can be effectively improved.With the in-depth development and implementation of the intelligent distribution network construction and transformation action plan,the collection cost of distribution network data is continuously reduced,the available data types are more diverse,and the concept of distribution network big data gradually emerges.Research on the integration of big data technology into the distribution network has become possible.Therefore,this thesis carries out the research on the forecast of the distribution network repair order based on big data technology,and applies the big data technology and artificial intelligence algorithm to the prediction of the distribution network repair order,which serves as the algorithm basis of the fault repair module in the monitoring and analysis tool in special period and provides strong support for the distribution network repair command work.Starting from this,the main contents of this thesis are as follows:Firstly,the data processing method and the change law analysis method of the repair work order based on big data technology are proposed.The power distribution information system and the weather forecast system are fully utilized,and the preliminary data research work is done to minimize the waste of existing data and improve the data utilization rate.Then,the thesis establishes the distributed forecasting method of rush repair order.A regional large power grid is divided into several subnetworks according to the division of administrative regions and the correlation of the change curve of rush repair work.Combined with seasonal type and temperature and humidity index,each subnet selects the appropriate forecast model of rush repair work order.On the premise of ensuring the prediction accuracy,the model can be simplified as much as possible to facilitate the programming and reduce the memory consumption.Finally,the distributed prediction method of distribution network rush repair order based on big data technology was successfully applied to the fault rush repair module in the monitoring and analysis tool of S province power grid in special period,mainly realizing the prediction function of distribution network rush repair order.With the help of the system platform,the prediction data of rush repair work orders in several cities in S province were selected to verify the accuracy and feasibility of the prediction method in this thesis.