A Comparative Study in Forecasting Power Generation in Kuwait using Various Artificial Intelligence Models

Abstract

Kuwait has a high demand for electricity towards indoor cooling and desalination of water, which it meets mainly through its fossil fuel reserves. The low energy tariffs and the carefree consumption have led to noticeably high per capita CO2 emissions, and it has faced around 2 degrees rise in temperatures over the decades. To meet its power requirements and combat emissions, investments towards the development of renewable sources of energy has taken the front seat, which require reliable forecasting techniques to lay the roadmap for future action. Though multiple sources are on offer in the energy market and for decades such a choice was absent, it is reasonable to go a step further to ensure that every household gets the optimum energy mix both in terms of minimal costs and minimal carbon emissions. Thus, power management units are incorporated on supply lines to dynamically manage the allocation of power from multiple sources when the supply and demand both are variable. This work discusses the popular algorithms associated with the forecasting and prediction of power generation in this field, and dynamic allocation of units to the consumers as the supply from these sources varies.