A Survey on Applications of Machine Learning for Optimal Power Flow
Fouad Hasan, Amin Kargarian, Ali Mohammadi
Abstract
Optimal power flow (OPF) is at the heart of many power system operation tools and market clearing processes. Several mathematical and heuristic approaches have been presented in literature to solve OPF. The recent flourish of machine learning (ML) algorithms and advancement of computational resources, with unforeseen data availability, has motivated the power system community to embrace ML. Although many papers are published on the applications of ML for solving various power system problems, in case of OPF, the same research orientation is still in its early days. This paper presents a survey of recent studies that have applied ML to solve OPF-related problems and provides readers with visions on potential research directions in this field. The surveyed literature is categorized according to the type of studied OPF problems.