Suppr超能文献

利用定量相显微镜评估变压器油质量。

Transformer oil quality evaluation using quantitative phase microscopy.

作者信息

Xing Xinyi, Zhu Lin, Chen Chao, Sun Nan, Yang Chuankai, Yan Keding, Xue Liang, Wang Shouyu

出版信息

Appl Opt. 2022 Jan 10;61(2):422-428. doi: 10.1364/AO.440583.

Abstract

Transformer oil used in oil-filled electrical power transformers aims at insulating, stopping arcing and corona discharge, and dissipating transformer heat. Transformer running inevitably induces molecule decomposition, thus leading to gases released into transformer oil. The released gases not only reduce the transformer oil's performance but also possibly induce transformer fault. To prevent catastrophic failure, approaches using, e.g., chromatography and spectroscopy, precisely measure dissolved gases to monitor transformer oil quality; however, many of these approaches still suffer from complicated operations, expensive costs, or slow speed. To solve these problems, we provide a new transformer oil quality evaluation method based on quantitative phase microscopy. Using our designed phase real-time microscopic camera (PhaseRMiC), under- and over-focus images of gas bubbles in transformer oil can be simultaneously captured during field of view scanning. Further, oil-to-gas-volume ratio can be computed after phase retrieval via solving the transport of intensity equation to evaluate transformer oil quality. Compared with traditionally and widely used approaches, this newly designed method can successfully distinguish transformer oil quality by only relying on rapid operations and low costs, thus delivering a new solution for transformer prognosis and diagnosis.

摘要

充油式电力变压器中使用的变压器油旨在起到绝缘作用、阻止电弧和电晕放电,并消散变压器产生的热量。变压器运行不可避免地会引起分子分解,从而导致气体释放到变压器油中。释放出的气体不仅会降低变压器油的性能,还可能引发变压器故障。为防止灾难性故障发生,诸如色谱法和光谱法等方法会精确测量溶解气体以监测变压器油的质量;然而,这些方法中的许多仍存在操作复杂、成本高昂或速度缓慢的问题。为解决这些问题,我们提供了一种基于定量相显微镜的新型变压器油质量评估方法。使用我们设计的相实时显微镜相机(PhaseRMiC),在视野扫描过程中可以同时捕获变压器油中气泡的欠焦和过焦图像。此外,通过求解强度传输方程进行相位恢复后,可以计算出油气体积比来评估变压器油的质量。与传统且广泛使用的方法相比,这种新设计的方法仅依靠快速操作和低成本就能成功区分变压器油的质量,从而为变压器的预后和诊断提供了一种新的解决方案。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验