FAHIM RAHMAN

Engineering final year major project was done on Transformer Incipient Fault Diagnosis using Machine Learning Classifiers. A brief summary on the project is as follows -" Power transformers are an integral and one of the most expensive components in power systems. Timely and accurate detection of transformer faults is essential for a reliable power system network. Dissolved Gas Analysis(DGA) is the most widely performed test performed for transformer health monitoring due to its easiness and applicability for online diagnosis. However, because each approach is only suited for particular situations, the conventional DGA methods frequently produce erroneous predictions of the incipient defect in transformer oil. The project focuses on exploring new method for the interpretation of DGA data to identify the insipient transformer faults with significantly improved accuracy, using data augmentation and boosting algorithms within machine learning."