| Summary: | 碩士 === 國立臺北科技大學 === 電機工程系研究所 === 99 === Oil-immersed power transformer needs to endure for an extended period of time high voltage, high current, temperature rise, and the impacts of electromagnetic and mechanical forces, and any malfunction may lead to large-scale power outages and other drastic losses. Oil-immersed power transformer is also highly expensive. It is therefore of crucial importance to keep the transformer in normal operation and to prolong its service life. Using effective maintenance mechanism and test-diagnostic techniques to measure the aging degree of a power transformer helps trace the current status of the power transformer and enables the operator to initiate proper responses to contingencies and irregularities.
The study accordingly aims at measuring the aging life of insulation paper by examining the elements and related parameters in transformer oil. The relationships between aging of insulation paper and major elements of transformer oil, notably furan, dissolved gases (CO2, CO), loading, hot-spot, temperature, moisture, and dielectric loss, are analyzed. Based on the standard and literature review, furan, moisture, and hot-spot temperature of oil help provide better diagnostic results. The study then proceeds to examine the methods frequently used to evaluate the aging life of transformer oil and compare the differences between the various methods. Obtained figures related to major affecting factors like furan, CO2, and CO are adopted to quantify the life loss of power transformer. Artificial neural network (ANN) is further applied to identify and examine the correlations between the aging of power transformer and affecting factors. Using two- and four-input-variable assessment models, the study has the functions of transformer life loss served as the ANN outputs to facilitate network training. The trained model is then used to measure and forecast the life loss of aging power transformers.
|
|---|
