Research On Data Polymerization Model Of Power Transformer Insulation On - Line Monitoring And Its Application

Transformer, the core in grid energy conversion and transmission, is a key hub device in the first defense line of power grid security. Mainly regular maintenance is used to assess a transformer’s property and operating status, but this method has several defects:it ignores the actual state of a transformer; the research into the trends and variation rules expressed in the detected information is not thorough enough; it fails to look at the problem from the perspective of "information-driven"; it pays no attention to the potential change of a transformer, thus meets difficulty in assessing transformer condition.Insulation devise, an important part in electrical equipment structure, has become a weak link along with the increase of voltage class in power system. When the insulation devise in electrical equipment is damaged and causes accidents, the electrical system will not operate stably, and the national economy will also suffer great loss. Nowadays the most widely-used insulation material in transformers is oil-impregnated paper insulator, but this kind of insulator will constantly age in operation under the impact of multiple factors, such as electricity, heat and machinery.This paper, through deep study of information aggregation architecture based on Internet of Things (IOT) and various information aggregation algorithms used relative deterioration degree to measure the insulation degradation of transformers, according to the insulation aging mechanism of transformers, to realize the division of a transformer’s insulation condition. This paper studied the kernel principal component analysis and the transformer insulation condition assessment method of Support Vector Machine(SVM), put forward the information aggregation model of transformer insulation condition assessment based on canonical correlation analysis, claimed to make decisions on transformer insulation according to the correlation analysis results to determine the insulation operating status of transformers, and to use case analysis to certify the effectiveness and superiority of the information aggregation model.The information aggregation model based on correlation analysis has a number of advantages and can be used to make an effective assessment of transformer insulation condition. It integrates local features into the global ones, obtains a better effect in extracting transformer insulation parameters, and realizes the function of intelligent processing of massive data. This model can be used to direct transformer operation maintenance and repairs, to prolong transformers’service life, and to achieve its economical and efficient operation.