Skip to Main Content
The quality of short term load forecasting can improve the efficiency of planning and operation of electric utilities. In this work, for determining the competitive learning model, the particle swarm optimization (PSO) technique is used as a training algorithm to adjust the weights of the artificial neural networks (ANNs) model to predict hourly loads. The feature of PSO is to fly potential solutions through hyperspace, accelerating toward better solutions. Thus the training phase should result in obtaining the weights configuration associated with the minimum output error. The historical load and weather information were trained and tested over a period of one season through two years. Generalized error estimation is done by using the reverse part of the data as a "test" set. The results were compared with conventional back-propagation algorithm and yielded encouraging results.