Oil immersed transformer fault diagnosis based on cross entropy algorithm optimized support vector machine. (October 2018)

• Record Type: Journal Article
• Authors: Bian, L; He, H
• Is Part Of: IOP conference series. Volume 188(2018)
• Publication Date: 2018-10
• Subjects: Earth sciences -- Periodicals; Environmental sciences -- Congresses; Environmental sciences -- Periodicals; 550.5
• DOI: 10.1088/1755-1315/188/1/012053
• Languages: English
• ISSNs: 1755-1307

Abstract

Abstract: With the increase of the voltage level of the power grid and the increase of its capacity, the probability of transformer failure is getting higher and higher. In order to discover the early latency faults of transformers, a cross-entropy algorithm was proposed to optimize the support vector machines. This method established a support vector machine classification model and used the cross-entropy algorithm to optimize the penalty factor and kernel function parameters. The transformer fault data is used to verify the classification model and compared with the test results of other algorithms. The results show that the accuracy of this algorithm for oil-immersed transformer fault diagnosis has reached 86.7%, which is higher than that of genetic algorithm and particle swarm algorithm, and iterating over 6 times of the fitness curve tends to be smooth and takes less time. After many tests, the forecast results are stable.