Regional load forecasting in Taiwan - applications of artificial neural networks

Up to now, the general style of load forecasting emphasized aggregate load forecasting. Such load forecasting results not only cannot identify where the power load takes place but also is not helpful for power facilities construction location planning. On the other hand, the power industry has been moving toward a deregulated environment recently. The results of regional load prediction could be used by power retailers to find their potential business opportunities. For transmission and distribution operators, accurate regional load forecasting can help them in long term power system planning and construction. Thus, regional load forecasting is getting more and more important for electricity providers in a deregulated power market. In this paper, empirical data are collected to formulate an artificial neural network model to predict the regional peak load of Taiwan. Based on the forecast results, some suggestions for Taiwan power market providers are presented. (C) 2002 Elsevier Science Ltd. All rights reserved.