PROJECT TITLE :

Adaptive neuro-fuzzy inference system approach for simultaneous diagnosis of the type and location of faults in power transformers

ABSTRACT :

Electrical, mechanical, and thermal stresses will degrade the quality of the insulation in power transformers, inflicting faults [one]. Many methods are used for fault diagnosis in transformers, e.g., dissolved gas analysis (DGA), measurement of breakdown voltage, and tan ??????, pollution, sludge, and interfacial tension tests [a pair of]. Of those, DGA is the foremost frequently used. Thermal and electrical stresses result in fracture of the insulating materials and the release of several gases. Analysis of these gases may offer information on the sort of fault. Numerous standards are instructed for the identification of transformer faults primarily based on the ratio of dissolved gases within the transformer oil, e.g., International Electrotechnical Commission (IEC) standards [three]?????????[7], and these standards has been quoted in many papers, e.g., [eight]?????????[15]. However, they are incomplete in the way that, in some cases, the fault can not be diagnosed or located accurately. Intelligent algorithms, e.g., wavelet networks [sixteen], neuro-fuzzy networks [17], [18], fuzzy logic [8], [twelve], and artificial neural networks (ANN) [two], [nine], [10], [19], [twenty] have been used to enhance the reliability of the diagnosis. In these algorithms, the sort of fault is diagnosed 1st, and therefore the fault is then located using the ratio of the concentrations of CO2 and CO dissolved in the transformer oil [twenty one], [22]. The algorithms don't seem to be entirely satisfactory. The wavelet network has high efficiency however low convergence, the fuzzy logic technique has a limited variety of inputs and, in some cases, it is terribly difficult to derive the logic rules, and therefore the ANN want reliable coaching patterns to enhance their fault diagnosis performance. During this paper, we present a new method for simultaneous diagnosis of fault sort and fault location. It uses an adaptive neuro- fuzzy inference system (ANFIS) [23]?????????[27], based mostly on DGA. The ANFIS is initial ?????????trained????????? in accordance with IEC 599 [three], so that it acquir- s some fault determination ability. The CO2/CO ratios are then thought-about extra input information, enabling simultaneous diagnosis of the sort and location of the fault. The results obtained by applying it to six transformers are presented and compared with the corresponding results obtained using ANN and some other standards and ways.