A proposed intelligent short-term load forecasting hybrid models of ANN, WNN and KF based on clustering techniques for smart grid

Smart grid is one of the most important topics to be covered with the increasing penetration of renewable energy in the power system grid to improve grid energy efficiency by managing the relationship between the demand and the generation. Load forecasting is playing a crucial role in this process as well as the output power generation from different renewable energy resources. The accuracy of the forecasting models is very important to deal with the new energy generation and consumption. Conventional approaches used in the literature work done for load forecasting can not handle the requirements of new generation of renewable energy and their uncertainties. This paper is proposing a novel technique for short-term load forecasting based on hybrid of different models and using clustering techniques to improve the overall system performance and quality. These models involve different combinations of Kalman filtering (KF), Wavelet and Artificial Neural Network (WNN and ANN) schemes. Six different models are proposed based on the clustering techniques. Simulations proved higher performance of the proposed models. The data used is commercial data, so it is scaled in this paper. The proposed work is validated by using different dataset for two different locations in Egypt and Canada.