Synthetic Electricity Consumption Data Generation Using Tabular Generative Adversarial Networks

Abstract

Generating synthetic electricity consumption data is crucial for developing efficient energy systems in smart cities. In this paper, we propose the use of Tabular Generative Adversarial Networks (Tabular GAN) for generating synthetic data for residential electricity consumption. Tabular GANs have been used in various domains and have shown promising results in generating high-quality synthetic data. The performance of our proposed method was evaluated by comparing the probability density, mean, standard deviation, and variances of the synthetic data with the original data. The results showed that the Tabular GAN method generated synthetic data that closely match the statistical characteristics of the original data and the simulation outcome indicated that the synthetic data generated by Tabular GAN could effectively simulate the patterns and behaviors observed in the original data. Overall, the proposed method demonstrates the effectiveness of using Tabular GANs for generating synthetic electricity consumption data.