Fault Diagnosis of Oil Immersed Transformer Based on the Support Vector Machine Optimized by Improved Fruit Fly Algorithm

Abstract The oil-immersed transformer is studied, and the support vector machine (SVM) algorithm is used. The radial basis is selected as the kernel function and is optimized by the improved fruit fly (IFF) algorithm based on the parameter characteristics to diagnose faults. By the simulation experiment, it is concluded that the proposed SVM algorithm optimized by using the IFF algorithm can not only avoid the local extremum problem but also show good generalization ability for small sample data processing, which has development potential in diagnosing power transformer faults.