A web based system for short-term forecasting of Algerian electricity load using artificial neural network

Abstract EN

The industry of electrical energy, in the majority of the developing countries, is changing from an operational approach centralized towards a competing one.

The notion of forecasting of electrical energy, by only one public service, is replaced by the concept that a competitive market is a mechanism more appropriate to ensure energy to consumers with a high reliability at less cost.

Algeria counts among these countries.

Under such competitive conditions, forecasting becomes a powerful tool in the search for higher outputs.

A valid model (or forecast tool) for the forecasts of load and / or electricity consumption, will help to reach this condition.

This work presents the results obtained by a prediction model based on Artificial Neural Network (ANN) and developed with Cwhich will facilitate the deployment on a server for a web based access.

This combination of web spread access by using C # programming language and neural network prediction abilities by implementing Multi Layer Perceptron (MLP) permits to use a web based neural network prediction system issued from these two technologies advantages.