Skip to Main Content
According to the non-linear relation characteristic of load, a short-term load forecasting model based on fuzzy neural networks was presented. In the model, fuzzy inference and defuzzification were completed by neural networks, and the neural networks weight values were given definite knowledge meaning. The membership function of fuzzy layer was selected to translate the input variables of load into fuzzy variables. Then a new inference algorithm was discussed to finish fuzzy inference. Finally, the forecasting load values were obtained by proper defuzzification. The simulation results show preferable forecasting capability of the model.