کنفرانس بین المللی ماشین و محرکه های الکتریکی

صفحه اصلی / The 4th International Conference on Electrical Machines and Drives

A Brief Review on the Application of Machine Learning for Transient Stability Prediction of Synchronous Generators to Proposing a Comparative Framework

نویسندگان :

Ali Abdalredha¹ Alireza Sobbouhi² ابوالفضل واحدی³

1- دانشگاه علم و صنعت ایران 2- دانشگاه شهید بهشتی 3- دانشگاه علم

کلمات کلیدی :

Data processing،Machine learning،Stability prediction،Transient stability

چکیده :

The accuracy and speed of real-time transient stability assessment (TSA) in power systems presents a significant challenge. Traditional methods often fall short in providing reliable and interpretable results. This paper proposes a novel framework utilizing machine learning techniques, including Decision Tree (DT), k-Nearest Neighbors (KNN), Support Vector Machine (SVM), and Logistic Regression (LR), to enhance the accuracy and reliability of TSA. This framework makes machine learning models more interpretable and useful in real-world scenarios by standardizing the test system, distributing the quantity of training and testing cases, choosing input data specifications carefully, and presenting TSA results with accuracy and remaining prediction time until instability. The promise of this approach for real-time applications in complex power networks is highlighted by the inclusion of real-world issues, such as noisy and missing data, and the emphasis on measures beyond correctness.