An algorithm for short term power data forecasting based on GM-ANN composite model

In urban power grid, the short-term power forecasting is mainly to forecast the power demand of different times in different time, which can be used to improve the power allocation and improve the efficiency of the city. The factors affecting the use of electricity are numerous, and can't be expressed by the exact mathematical model. Taking into account the advantages of gray prediction and neural network to nonlinear and fuzzy information processing capacity, the use of gray model GM (1,1) for power data modeling, the use of artificial neural network ANN on the actual value and the prediction of the residual model, and finally the above two models, GM-ANN combination of power forecasting model.