Quantum Machine Learning for Power System Stability Assessment

Transient stability assessment (TSA), a cornerstone for resilient operations of today's interconnected power grids, is a grand challenge yet to be addressed since the genesis of electric power systems. This paper is a confluence of quantum computing, data science and machine learning to potentially resolve the aforementioned challenge caused by high dimensionality, non-linearity and uncertainty. We devise a quantum TSA (qTSA) method, a low-depth, high expressibility quantum neural network, to enable scalable and efficient data-driven transient stability prediction for bulk power systems. qTSA renders the intractable TSA straightforward and effortless in the Hilbert space, and provides rich information that enables unprecedentedly resilient and secure power system operations. Extensive experiments on quantum simulators and real quantum computers verify the accuracy, noise-resilience, scalability and universality of qTSA. qTSA underpins a solid foundation of a quantum-enabled, ultra-resilient power grid which will benefit the people as well as various commercial and industrial sectors.