An Intelligent Adversarial Deep Forecasting Model for Load Demand Using Hybrid Modified DA-GAN

Extreme weather conditions can have a significant impact on the electricity load demand and energy management programs and thus can cause unexpected blackouts in the systems. To predict electricity load demand in extreme weather conditions, it is important to consider the different weather historical data in the analysis. This paper proposes an intelligent adversarial model for prediction of the consumers' electric load demand in the extreme weather condition. By analyzing past extreme weather events and the associated electricity load demand, a new predictive deep learning model developed that can be used to estimate electricity load demand in the future extreme weather conditions. The proposed model is constructed based on the generative adversarial network (GAN) and dragonfly algorithm (DA) to make a precise prediction. The generator network is trained to produce data that are similar to the extreme weather data, while the discriminator is trained to correctly classify real data from generated ones. A new modified DA is suggested to enhance the GAN training through an iterative process. The real dataset of weather and load in California over the years 2015–2020 are used to examine the accuracy of the proposed model.