Design Of Power Transformer Condition Assessment System

Transformer condition assessment as the basis and foundation of the state of maintenance, whether directly reflect the stable operation of the transformer can be healthy, regardless of any fault within the transformer corresponding parameters are beyond the normal range, but the results of these parameters or parameters are an important basis for transformer condition assessment. Therefore, in the evaluation process, the transformer unusual exhibit different degrees of state, these states are divided into different levels to assess the state of the transformer, the transformer condition in real time is the outcome of the assessm nt of the health status, it is also an important reference value in late troubleshoot. Therefore, this article is based on the experimental data for each transformer starting using fuzzy math to determine the relative degree of degradation of another generation of each parameter into its corresponding membership functions, for the amount of oil in the gas state, due to its own characteristics and importance, the results of three different levels according to the ratio of discharge failure or overheating is divided into different states by using three-ratio method, and referring the table directly. Finally, for the system to be able to achieve a more perfect self-learning function, using the membership function of the neural network is trained to make a preliminary assessment of the value and the expected close, trying to reduce the error in order to make the system more accurately on the same level in the latter part of the assessment of the transformer.