In order to accurately identify abnormal electricity data and make real-time correction to ensure the safety of users'e-lectricity consumption,a corrected model of abnormal electricity data of users in the power market with high participation of new energy users is proposed.Taking a province and city with a high proportion of new energy as an example,this paper analy-zes the possible abnormal data phenomenon of user electricity in its power market,builds a probability prediction model,identi-fies the abnormal data of user electricity through off-line training and online identification,combines with the threshold and e-lectricity prediction interval,and divides the identified abnormal data of user electricity into two categories according to the in-complete load data and no load data of users.It then obtains the abnormal data of user electricity that need to be corrected in the two types.It is taken as input,a corrected model of abnormal data of user electricity is built based on deep neural network,and the network outputs the correction results of abnormal data of user electricity.The experimental results show that the model can identify the abnormal data of users'electricity consumption and accurately identify the abnormal time of electricity con-sumption data,and can correct the identified abnormal data of electricity consumption in real time.
关键词
高比例新能源/电力市场/用户电量/异常数据/修正模型/深度神经网络
Key words
high proportion of new energy/power market/user electricity/abnormal data/corrected model/deep neural net-work
维普
万方数据