Power System Operation Stability Assessment Method Based on Deep Convolutional Neural Network

INTRODUCTION:  For the assessment of power system stability, a power system assessment method based on a deep convolutional neural network is studied. OBJECTIVES: Through the improvement of the integrated convolutional neural network (CNN) network model, the impact of insufficient transient stabil...

Full description

Saved in:
Bibliographic Details
Main Authors: Jinman Luo, Yuqing Li, Qile Wang, Liyuan Liu
Format: Article
Language:English
Published: European Alliance for Innovation (EAI) 2025-05-01
Series:EAI Endorsed Transactions on Energy Web
Subjects:
Online Access:https://publications.eai.eu/index.php/ew/article/view/7572
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:INTRODUCTION:  For the assessment of power system stability, a power system assessment method based on a deep convolutional neural network is studied. OBJECTIVES: Through the improvement of the integrated convolutional neural network (CNN) network model, the impact of insufficient transient stability assessment caused by sample misjudgment and sample omission is effectively reduced. METHODS: We adopt the hierarchical real-time prediction model to evaluate the stability and instability of the determined stable samples and unstable samples, thereby improving the timeliness and accuracy of transient evaluation. RESULTS: Through experimental comparison, the integrated CNN network model in this study has obvious advantages in accuracy compared with the single CNN network. Compared with other algorithm reference models, this model has a higher evaluation accuracy of 98.39%, far exceeding other comparison models. CONCLUSION: By further evaluating the model’s accuracy, it is proved that the model can provide an effective reference for the follow-up power system stability prevention and has important application value.
ISSN:2032-944X