Generation of Synthetic Multi-Resolution Time Series Load Data

TL;DR

This paper presents LoadGAN, an end-to-end generative framework that creates realistic synthetic multi-resolution time series load data for power systems, aiding research with limited real datasets.

Contribution

It introduces a novel combination of PCA and GANs to generate diverse, multi-resolution load data, and provides an open-source tool for easy access and use.

Findings

Successfully captures diverse load characteristics

Generates data at multiple sampling rates and durations

Open-source tool facilitates research and data sharing

Abstract

The availability of large datasets is crucial for the development of new power system applications and tools; unfortunately, very few are publicly and freely available. We designed an end-to-end generative framework for the creation of synthetic bus-level time-series load data for transmission networks. The model is trained on a real dataset of over 70 Terabytes of synchrophasor measurements spanning multiple years. Leveraging a combination of principal component analysis and conditional generative adversarial network models, the scheme we developed allows for the generation of data at varying sampling rates (up to a maximum of 30 samples per second) and ranging in length from seconds to years. The generative models are tested extensively to verify that they correctly capture the diverse characteristics of real loads. Finally, we develop an open-source tool called LoadGAN which gives researchers access to the fully trained generative models via a graphical interface.