A Hybrid Method Based on Discrete Wavelet Transform for Load Forecasting by Using a Two Dimensional Model

Forecasting the electric energy demand and identification of the trend of changes are the effective key factors in the planning and operation of the power network. The knowledge about electrical energy consumption is the basis for planning and deciding in power systems. In this paper by introducing a combination of wavelet transform and least error square method and using a two-dimensional load model, monthly peak demand of Zanjan grid will be forecast in the long-term horizon. Then, to evaluate the effectiveness of this method, the results of applying the proposed hybrid method is compared with several methods. The evaluation results indicate that the proposed hybrid approach has a higher capability in load forecasting. Finally, due to the significant improvement in prediction performance evaluation index, this method will be used to predict future demand of Zanjan province.