A new forecasting framework for volatile behavior in net electricity consumption: A case study in Turkey

Electricity is a significant form of energy that cannot be stored physically and is usually generated as needed. In most research studies, the main aim is to ensure that sufficient electricity is generated to meet future needs. In order to avoid waste or shortage, a good system needs to be designed to constantly maintain the level of electricity needed. It is necessary to estimate independent factors because future electricity volume is based not only on current net consumption but also on independent factors. In this paper, a new framework is proposed to first estimate future independent factors using SARIMA (seasonal auto-regressive iterative moving average) method and NARANN (nonlinear autoregressive artificial neural network) method, both of which are called a "forecasted scenario approach". Subsequently, based on these scenarios, a LADES (LASSO-based adaptive evolutionary simulated annealing) model and a RADES (ridge-based adaptive evolutionary simulated annealing) model are applied to forecast the future NEC (net electricity consumption). The proposed approaches are then validated with a case study in Turkey. The experimental results show that our approach outperforms others when compared to previous approaches. Finally, the results show that the NEC can be modeled, and it can be used to predict the future NEC.