Peak load demand forecasting by using a fuzzy classification system combined with the convolutive aggregation method

摘要

Throughout history a lot of research and load models have been put forward to forecast the electric power demand in MV distribution and transport systems. Many different kinds of criteria are used in models for estimating electrical demand: direct or indirect models, determinist or stochastic models, short or long-term models HV, MV and LV models. Iberdrola has currently tackled the matter of LV models, within the new supply and planning field. In cooperation with Iberdrola, the Electric Technology Institute, ITE, has developed a model for estimating the load on three-phase LV lines and MV/LV transformers. The innovation of this model is the classification of customers associated with types of probability density function in quarter-hourly consumption. The model developed is of special use in planning and running distribution systems with disperse generation, where the forecast in estimation of demand is vital for control with a highly changing power flow.