Positive Active Power Missing Value Repairing with LSTM Prediction

Positive active power is one of the two major indicators for checking the power quality of the power grid. The missing of its value will seriously affect the results of fault diagnosis. This paper proposes a method for repairing missing positive active power data based on the LSTM prediction model. The identified normal historical positive active power is used as the input variable, and the positive active power data of the time node where the missing value is located is used as the output variable to iteratively predict the positive active power data, and modify the missing value according to the results. After testing, this method has good theoretical and practical application value in repairing the missing of positive active power data.