ACCEPTED FOR PRESENTATION IN 11TH BULK POWER SYSTEMS DYNAMICS AND CONTROL SYMPOSIUM (IREP 2022), JULY 25-30, 2022, BANFF, CANADA

The increasing deployment of end use power resources in distribution systems has led to the emergence of active distribution systems, which display significant voltage and loading variations throughout the day due to the variable nature of renewable energy resources. This variability necessitates effective control to ensure reliable power supply under standard voltage and frequency conditions. Traditional optimization methods struggle due to the complexity and dimensionality of the problem, making global optimization challenging for the vast array of small resources involved. In this paper, we explore the potential of Artificial Intelligence (AI) methods, specifically neural networks enhanced with self-attention mechanisms, as a viable alternative for system optimization. We present our approach, detailing its application to active distribution systems, and provide preliminary results that suggest significant promise in utilizing these AI techniques for improved power system control. Our findings indicate that these advanced methods can effectively address the complexities associated with modern distribution systems, paving the way for more sustainable energy management solutions.