Power transformer fault diagnosis and condition monitoring using hybrid TDO-SNN technique

In this manuscript, a hybrid approach for optimal detection and classification of fault on load tap changer of power transformer is presented. The proposed hybrid approach is the Tasmanian Devil Optimization, Spike Neural Network and commonly called to as the TDO-SNN method. The major objective of the proposed approach is to minimizing the error and enhances the accuracy, safety, and efficiency of power transformer operation within electrical power systems. Classification of transformer faults is applied in two phases with the ultimate goal of SNN detection. In various situations, thenormalSNN first phase is used to detect the healthy or unhealthy state of the transformer. The second phase SNN process uses TDO from the perspective of minimum error objective function. Classifying the transformers' ill state in order to identify the proper faults for protection is the 2nd stage of the SNN. At the first stage, the TDO-SNN method plays an estimate process to protect the transformer and detect the fault in the transformer. The TDO-SNN technique reduces the problem of transformer fault deduction and classification and the accuracy of the system is high. Then, the model is executed in MATLAB platform and the implementation is designed with current procedures.