Expert system for power transformer diagnosis

Power transformers are the most critical part of power electrical system. The oil and the insulation system are subjected to degradation for many chemicals inside them, they are the result of an initial problem that can be predicted. In this research, the intelligent diagnosis system based on component analysis (PCA) and adaptive decision system based on fuzzy logic permits to realize a dissolved gas analysis (DGA) [1] to predict incipient fault diagnosis by different methods, to obtain deterioration rates and health index, besides it allows to analyze the degree of polymerization (DP) for the remaining life of the equipment. It determinates the comparative rates and determinate the efficiency and effective of this diagnosis system, it improves the international standards IEEE, IEC, among others. The correct diagnosis performance of the PCA and fuzzy logic is calculated on 107 samples on Peruvian power electrical systems with excellent results.