A Neuro-fuzzy Approach for Predicting Load Peak Profile

Abdellah Draidi (Laboratoire de genie électrique de Constantine, Departement of electrical engineering, University of Constantine 1.)
Djamel Labed (Laboratoire de genie électrique de Constantine, Departement of electrical engineering, University of Constantine 1.)

International Journal of Electrical and Computer Engineering
Vol 5, No 6: December 2015

Load forecasting has many applications for power systems, including energy purchasing and generation, load switching, contract evaluation, and infrastructure development. Load forecasting is a complex mathematical process characterized by random data and a multitude of input variables.To solve load forecasting, two different approaches are used, the traditional and the intelligent one.Intelligent systems have proved their efficiency in load forecasting domain. Adaptive neuro-fuzzy inference systems (ANFIS) are a combination of two intelligent techniques where we can get neural networks and fuzzy logics advantages simultaneously. In this paper, we will forecast night load peak of Algerian power system using multivariate input adaptive neuro-fuzzy inference system (ANFIS) introducing the effect of the temperature and type of the day as input variables.