| Summary: | One of the most crucial parameters in operating a vacuum interrupter (VI) is internal pressure. The failure of switching or insulation occurs when the pressure rises above a specific level. Characteristics of partial discharge (PD) in VI can be used to measure the internal pressures of VI. This paper defines a classification problem for the degree of internal pressure in VI using PDs, which were measured using a capacitive PD coupler. Then, we propose a deep neural network to monitor the internal pressure of VI by analyzing PDs. Experimental results show that the proposed deep neural network monitors the internal pressure range, from <inline-formula> <tex-math notation="LaTeX">$1.0 \times 10^{-2}$ </tex-math></inline-formula> torr to 10 torr in VI. The classification performance of the proposed method is significantly better than those of machine learning algorithms such as support vector machines and <inline-formula> <tex-math notation="LaTeX">$k$ </tex-math></inline-formula>-nearest neighbor algorithm and the proposed method achieves an 100% classification accuracy.
|
|---|