An fuzzy forecasting algorithm for short term electricity loads based on partial clustering

Abstract

A short-term load forecasting method using fuzzy inference based on partial clustering has been proposed. In this forecasting system, local feature selection is first used to find the proper feature subset for regional load forecasting. Then, the typical regional load patterns are obtained by partial feature clustering. Lastly, independent regional load forecasting models are established by using fuzzy inference based on these local typical load patterns. The performance of this method is proved to be satisfactory by a case study in Jiangning city of China.