کنفرانس مهندسی برق ایران

صفحه اصلی / سی و چهارمین کنفرانس بین المللی مهندسی برق

Fast Optimal AC Power Flow by Machine Learning

نویسندگان :

Reza Mohammad Yousefi¹ Ahmad Salehi Dobakhshari²

1- دانشگاه گیلان 2- دانشگاه گیلان

کلمات کلیدی :

Deep Learning،Frequency Domain،Machine Learning،Optimal Power Flow،Power System Optimization

چکیده :

The integration of renewable energy sources and flexible loads into modern power systems calls for solving the Optimal Power Flow (OPF) problem more frequently. While numerous machine learning-based methods have been explored for fast OPF solutions, they often perform poorly in large-scale systems with more than thousands of buses. This paper presents a novel machine learning (ML) method, which operates in the frequency domain to capture long-range dependencies inherent in power flow dynamics. This ML architecture provides an effective solution to the AC-OPF problem, offering high accuracy and efficiency in power flow predictions. Simulations on IEEE test cases demonstrate the superior generalization of the proposed ML method to unseen load profiles, showcasing its potential for real-time optimization in renewable-integrated power systems.