O.A. Alsayegh, O.A. Al-Matar, F.A. Fairouz, and A. Al-Mulla Ali (Kuwait)
Artificial neural networks, Kuwait, long-term forecasting, simulation model, socio-economic factors
In this paper, the electric peak power demand in Kuwait up to year 2025 is predicted using artificial neural network (ANN) based simulation model. The contribution of this work is the inclusion of the analysis of the effect of the air-conditioning (A/C) units on the long-term power demand. In addition to the A/C factor input, four socio economic factors are selected as inputs to the simulation model: gross national product (GNP), population, number of buildings, and index of industrial production (IIP). Socio-economic data from 1978 to 2000 years are used for training the ANN. Analysis result shows that the average peak power increase rate is 4100 MW per 5 years. Various scenarios were simulated to estimate the long-term electric peak power demand. Such scenarios include variation of population, number of buildings and quantity of A/C units import. The residual error that is produced as a result of the long-term forecasting is estimated.
Important Links:
Go Back
