Evaluation And Prediction Of Distribution Network Operation Status Based On Multi-source Data Fusion

At present,there are many types of distribution network data,complex structure,heterogeneous data,and data from different data acquisition systems and terminal characteristics.At this stage,the operation status of distribution network lines and power load forecasting have the characteristics of inaccuracy and hysteresis,which seriously affects the operation status of distribution network lines and the effectiveness and timeliness of distribution network load,and restricts the further development of distribution network technology in China.Based on this,this paper is combined with the actual situation of the distribution network in Leshan area,that is,the power supply company has various software systems to collect the basic data and operation data of the distribution network,and adopts the idea of combining big data technology to propose a distribution network line operation state evaluation and distribution network load forecasting method.The main research contents are as follows :(1)Aiming at the problem that the distribution network has many kinds of data,complex structure,heterogeneous data,and data from different data acquisition systems and terminals,a data processing method based on down-sampling and principal component analysis is proposed.Data cleaning and down-sampling are used to reconstruct data to improve sparsity and balance sample classes.On this basis,principal component analysis is used to determine the main components to solve the problem of data complexity and uncertainty of the main influencing factors,so as to realize the reconstruction of distribution network industry data and the reduction of sample number.(2)Aiming at the problem of hysteresis and inaccuracy in the evaluation of distribution network line operation state data,a distribution network line operation state evaluation method based on convolutional neural network is proposed.The distribution network state is divided into normal state,attention state,abnormal state and serious state.(3)Aiming at the problem of load change in the power consumption process of the distribution network,according to the historical load power consumption data,a distribution network load forecasting method based on long-term and short-term memory neural network is proposed.Considering the influencing factors of the distribution network,the medium-term load of Leshan City is predicted,and the average absolute percentage error and root mean square error are used as evaluation indicators to evaluate the prediction results.