A novel measurement-based procedure for dynamic equivalents of electric power systems for stability studies using improved sine cosine algorithm

Abstract

Dynamic equivalent (DE) is an important process of multi-area interconnected power systems. It allows to perform stability assessment of a specific area (area of interest) at minimum cost. This study is intended to investigate the dynamic equivalent of two relatively large power systems. The fourth-order model of synchronous generators with a simplified excitation system is used as equivalent to the group of generators in the external system. To improve the accuracy of the estimated model, the identification is carried in two stages. First, using the global search Sine Cosine Algorithm (SCA) to find a starting set values, then this set is used as starting point for the fine-tuning made through the Pattern Search (PS) algorithm. To increase the reliability of the model’s parameters, two disturbances are used to avoid the identification based on a specific event. The developed program is applied on two standard power systems, namely, the New England (NE) system and the Northeast Power Coordinating Council (NPCC) system. Simulation results confirm the ability of the optimized model to preserve the main dynamic properties of the original system with accuracy.

Enhanced SVM Algorithm-Based Dynamic Early Warning System for College English Ideological and Political Course Education Using Machine Learning

“…Several studies, like Benmiloud and Arif (2021), have acknowledged the transformative role of machine learning in education. While these works discuss the broader implications and potential benefits of applying machine learning to teaching and learning processes, they do not delve into the development of a specialized early warning system for ideological and political education within college English courses.…”