On line monitoring method for overheating fault of oil immersed transformer based on oil’s chromatographic data

As an important component of the smart grid, oil immersed transformers can ensure the stability of the grid through pre testing, monitoring, and diagnosis. Therefore, an online monitoring method for overheating faults of oil immersed transformers based on oil chromatography data is proposed. The article first analyzes the common types of faults in oil immersed transformers, analyzes the gas production mechanism of transformer oil, then constructs an oil chromatography data acquisition model under fault conditions, completes oil chromatography signal processing, and finally extracts oil chromatography data features. Based on support vector machine, the overheating fault diagnosis of oil immersed transformers is completed. Taking the DC resistance test data of 127 transformers in a certain region over the past 9 years as an example, the experimental results are as follows: using the proposed method, the overheating fault detection effect of oil immersed transformers is good, with a false alarm rate of about 0.16 %, a false alarm rate controlled at about 3.95 %, a fault diagnosis accuracy controlled at about 3.95 %, an average response time of only 1.39 s, and an average cost of 12.57 MBit, which is better than the comparative method and has a better application effect.