PREVOS-DZ: A Short-Mid Term Algerian Electric Load Forecasting Software

Abstract

Electricity demand is nowadays a very important development factor for growing economies. In order to support daily electrical demand and meet the customer daily needs, electricity and companies operators must rely on powerful and precise forecasting tools. PREVEOS-DZ was designed and implemented in order to predict Short and Midterm electrical energy and power demand in order to plan day to day production planning as well as yearly electrical development established by the Algerian Operator System. The software is based, on the multiple regression approach as well as Artificial Neural Networks, especially for power forecasting, offering models in both paradigms. The Forecast strategy is implemented not only for global national forecast but also for regional demands, even if the electrical grid is fully connected. PREVOS-DZ is implemented using C # and allows full compatibility with Excel tabular, used mostly by the System Operator. The results of PREVELOS-DZ obtained as a prevision exercise were validated on data from 2018 and presented in this paper.