Extended Design of Data Return and Equipment Remote Control for SF6 Gas Leak Detectors

• Published in: 2024 3rd International Conference on Energy and Electrical Power Systems (ICEEPS) pp. 989 - 994
• Main Authors: Lai, Junju, Liu, Wei, Huang, Liehong, Wang, Wei, Huang, Shihua, Guo, Zhijun
• Format: Conference Proceeding
• Language: English
• Published: IEEE 14.07.2024
• Subjects: Chemical industry; convolutional neural network (CNN); Convolutional neural networks; gas leak detection; Infrared imaging; Monitoring; Protection; Remote control; Robustness; Sensitivity; Substations; Sulfur hexafluoride; Sulfur hexafluoride (SF6)
• DOI: 10.1109/ICEEPS62542.2024.10693152

With the wide application of sulfur hexafluoride (SF6) gas in high-voltage power equipment, its potential leakage risks and environmental impacts have attracted widespread attention. Traditional monitoring methods have limitations in detection sensitivity and spatial coverage capability, while the development of artificial intelligence technology provides new possibilities for SF6 leakage detection. In this paper, we address the shortcomings of the existing monitoring methods and combine infrared imaging technology and convolutional neural network (CNN) to propose a new design for data return and equipment remote control extension of SF6 gas leak detection device. The design captures the thermal image of gas leakage through infrared imaging and processes the image data using CNN in order to realize the automatic detection of SF6 gas leakage. The experimental results show that the system is able to adapt to different light and background conditions and has good robustness, which provides a timely and all-weather supervision means for the power and chemical industry, and effectively improves the safety and environmental protection of the power system.