DefinePK

Abstract

Load forecasting have been a significant area of research as well as it has been a critical problem in planning and operation of electric power generation. In order to predict electrical load, the technique load forecasting is immensely used in global world. In addition, for power system planning the integration of short term load forecasting has been significant. Based on electric generating company, it is significant for them to analyze the market load demand in order to produce accurate power specifically in the deregulated market. This research paper have used half-hourly load data that have been gathered for short-term load forecasting in order to develop an accurate model for New South Wales. The use of MATLAB tool box with the integration of multilayer feed forward neural network has been used. The training has been provided with the use of Levenberg-Marquardt back propagation in order to evaluate the result as well as performance of the model during testing, training and validation. The effectiveness of Mean Absolute Percentage Error has been also considered in the paper. The result illustrated that this method is highly accurate and simple with minimum error as well as can be used for short term load forecasting.