Semi-Supervised Multi-Task Learning Based Framework for Power System Security Assessment

TL;DR

This work tackles the challenge of dynamic power-system security assessment under topology changes and renewable variability by introducing a Semi-Supervised Multi-Task Learning framework that fuses a conditional masked auto-encoder with a supervised classifier. It adds topology awareness through a Singular Value Sequence–based topological similarity index and a Mahalanobis-distance confidence measure to quantify prediction reliability, while leveraging time-domain simulations and infeasibility certificates to generate scalable training data. Empirical results on the IEEE 68-bus system show that SS-MTL outperforms baseline methods in accuracy and robustness, including under bad-data conditions, and achieves around 200x speedup over traditional time-domain analysis. The approach offers a practical, scalable screening tool with open datasets and code to promote reproducibility and industrial adoption, enabling operators to rapidly assess grid security under diverse contingencies and topologies.

Abstract

This paper develops a novel machine learning-based framework using Semi-Supervised Multi-Task Learning (SS-MTL) for power system dynamic security assessment that is accurate, reliable, and aware of topological changes. The learning algorithm underlying the proposed framework integrates conditional masked encoders and employs multi-task learning for classification-aware feature representation, which improves the accuracy and scalability to larger systems. Additionally, this framework incorporates a confidence measure for its predictions, enhancing its reliability and interpretability. A topological similarity index has also been incorporated to add topological awareness to the framework. Various experiments on the IEEE 68-bus system were conducted to validate the proposed method, employing two distinct database generation techniques to generate the required data to train the machine learning algorithm. The results demonstrate that our algorithm outperforms existing state-of-the-art machine learning based techniques for security assessment in terms of accuracy and robustness. Finally, our work underscores the value of employing auto-encoders for security assessment, highlighting improvements in accuracy, reliability, and robustness. All datasets and codes used have been made publicly available to ensure reproducibility and transparency.

Figures (7)

• Figure 1: The proposed Machine Learning framework for Dynamic Security Assessment (DSA)
• Figure 2: Process flow diagram of time domain simulations to create a database
• Figure 3: Machine Learning model training stage
• Figure 4: Multi-task Semi-supervised Conditional masked Deep Auto-encoder
• Figure 5: Performance of the algorithms on the IEEE 68 bus system