Application of Neural Network-based Combining Forecasting Model Optimized by Ant Colony In Power Load Forecasting

For non-linear and gray of power load forecasting, this paper proposed a new combining forecasting model. First optimize the parameters of the GM(1, 1, ¿) forecasting model with ant colony algorithm, and predict a set of load values; then predict another set of load values with Auto-regressive integrated moving average model (ARIMA). The forecasting results of ant colony gray model and ARIMA model were put as the input of RBF neural network to be forecast and trained. Therefore, an RBF neural network-based combining forecasting model was built. The results show that the combining model combines the advantages of different methods, and greatly improves the accuracy of load forecasting.