Oil-immersed Transformer Internal Thermoelectric Potential Fault Diagnosis Based on Decision-tree of KNIME Platform

The safety of power system is always affected by operating state of oil-immersed transformer directly. In order to improve reliability of power supply system, it is important to find and solve internal thermoelectric potential faults as soon as possible. In the past years, various of effective fault diagnosis techniques have been proposed, including chromatographic analysis, fault characteristic gases, IEC three-ratio method. Recently, artificial intelligence technique has been extensively used, such as neural network, genetic algorithm and so on. Although these methods have been applied in smart grid system effectively, we need a method with more accurate and more efficiency, visualization, and intelligence. So we proposed a method to solve internal thermoelectric potential fault diagnosis in this paper based on decision-tree C4.5 algorithm of KNIME platform. The experimental data from dissolved gas analysis (DGA) is used to illustrate performance of proposed decision-tree model. This designed model will obtain right results in short time by building a workflow in KNIME platform and convenient parameter setting which can ensure the normal operation of transformers.