Imbalanced Sample Generation and Evaluation for Power System Transient Stability Using CTGAN

Although deep learning has achieved impressive advances in transient stability assessment of power systems, the insufficient and imbalanced samples still trap training effect data-driven methods. This paper proposes a controllable sample generation framework based on Conditional Tabular Generative Adversarial Network (CTGAN) to generate specified samples. To fit complex feature distribution samples, proposed firstly models as tabular data uses Gaussian mixture normalize data. Then we transform multiple conditions into single conditional vector enable multi-conditional generation. Furthermore, this introduces three evaluation metrics verify quality generated framework. Experimental results IEEE 39-bus system show that effectively balances significantly improves performance models.