Title: Digital twin-based fault detection for intelligent power production lines

Authors: You Zhou; Xuefeng Qian; Dan Xu; Can Zhao; Kejun Qian

Addresses: State Grid Suzhou Power Supply Company of State Grid Jiangsu Electric Power Co., Ltd., Suzhou Jiangsu, 215004, China ' State Grid Suzhou Power Supply Company of State Grid Jiangsu Electric Power Co., Ltd., Suzhou Jiangsu, 215004, China ' State Grid Suzhou Power Supply Company of State Grid Jiangsu Electric Power Co., Ltd., Suzhou Jiangsu, 215004, China ' State Grid Suzhou Power Supply Company of State Grid Jiangsu Electric Power Co., Ltd., Suzhou Jiangsu, 215004, China ' State Grid Suzhou Power Supply Company of State Grid Jiangsu Electric Power Co., Ltd., Suzhou Jiangsu, 215004, China

Abstract: Digital twin technology realises real-time capture of system operation status, real-time monitoring and prediction of potential risks. In view of this, a fault detection method based on digital twin of power production line is proposed, where the attention technology required by virtual production line fault capture technology and model establishment combined with deep reinforcement learning model is used to analyse the power production line and realise fault detection. The method takes the node-related features of the visualisation equipment and power production line as input, analyses the possible production line faults through computer vision, and performs image recognition on all the collected pictures. The feature data collected by installing inspection equipment has rich information and spatiotemporal accompanying information of intelligent power production line, and the fault detection model of intelligent power production line constructed by digital twin has high confidence. The experimental results verified the effectiveness of the proposed method.

Keywords: digital twin; intelligent production line; deep learning; attention technology; power production line fault detection.

DOI: 10.1504/IJCSE.2024.139719

International Journal of Computational Science and Engineering, 2024 Vol.27 No.4, pp.385 - 392

Received: 21 Feb 2023
Accepted: 06 Jun 2023

Published online: 05 Jul 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article