Identification and Management to Anomalous Data in Short-term Load Forecasting
Chen Ya-hong · 2002
The influence of anomalous data consists in the historic load data to load forecasting is analysed .Different management methods on the basis of engineering reality are given to various anomalous data.Especially,an arithematic of supplementing load data after the forecast time of the day during forecasting day is introduced with its efficiency proved by an example.The example shows the change of forecasting errors before/after dealing with the anomalous data at the same time.