Reliability Analysis Of Lulong County Distribution Network Based On Big Data Technology

With the rapid development of distribution network informatization and the increasing influence of power demand,the amount of data in distribution systems is increasing.Traditional reliability analysis methods become no longer applicable because of the narrow data access channels and weak data integration processing capabilities,etc.The size and characteristics of distribution network data are in line with the big data features,which provides a new idea for distribution networks to analyze the reliability.This article propose a new method based on the big data to analyze the distribution networks reliability that utilizes large-scale data acquisition,storage,analysis,correlation,and forecasting technologies,mining from massive data,extracting key information,and predicting the parameters of reliability indicators at a certain time in the future to judge the operational reliability of the distribution network.The contents are as follows:Firstly,the characteristics of power failure,responsibility for power failures,and faulty apparatus in Lulong county during 2012 to 2016 were analyzed,and the proportion of each sub-item was clarified.Then,based on the five categories including regions,quantities of power failure households,types of faulty equipment and reasons for power failure,a two-step cluster analysis was performed on the data of power failure.The result of calculation confirmed that the cluster analysis was effective and successfully to extracted the statistical characteristics of power failure data in regions of Lulong county,and it is also feasible to mine out the evaluation indicators which is closely related to the grid reliability.Then,a four-dimensional index system is established for operational reliability of distribution networks that utilizes the data source,apply the principal component analysis method to mine the main evaluation indicators;based on the main indicators obtained to analyze the influencing factors and obtained the main influencing factors using the parallel association rule model.Taking the above-mentioned major influencing factors as input for forecasting and main evaluation indicators as output,using the artificial neural network to predict the operational reliability index value of distribution network in a certain time scale in the future.Finally,a 10 kV distribution system in Lulong County was taken as an example,the evaluation results are consistent with actual situation,which verifies the effectiveness of the proposed method.For the analysis results with lower reliability,combined with the current situation of distribution system in Lulong County put forward corresponding measures and suggestions.