loading page

Research on Transformer Fault Diagnosis with the application of Improved Manta Ray Foraging Optimization Algorithm optimizing Back-Propagation Network based on Orthogonal Experiment
  • YING ZHANG,
  • Longwu XU,
  • Chujie FENG
YING ZHANG
Guizhou University, Institute of Electrical Engineering Guizhou Guiyang, CN 550025

Corresponding Author:[email protected]

Author Profile
Longwu XU
Guizhou University
Author Profile
Chujie FENG
Electrical Power Research Institute of Guizhou Power Grid ,CO Ltd
Author Profile

Abstract

Oil-immersed transformers will inevitably have various latent faults during the ageing process, therefore, it is very important to correctly diagnose the status of the transformer in a timely manner. The traditional three-ratio method based on dissolved gas analysis (DGA) data has an insufficient coding problem, limiting the effect of fault diagnosis. To solve this problem, an improved manta ray foraging optimization (MRFO) algorithm is proposed to optimize the fault diagnosis model of the Back-Propagation (BP) Network in this paper. This improved algorithm primarily uses a multi-stage algorithm that combines logical mapping and opposition based learning (OBL) to provide the initial position for the MRFO to strengthen the global optimization ability of the algorithm. Meanwhile, three foraging strategies of modified MRFO algorithm by using the orthogonal experiment method are proposed, to adjust the exploration and development of the manta ray individual, so as to strengthen the optimization ability of the algorithm on specific problems. As the result, the optimal solution obtained by the improved MRFO algorithm is given to the weights and biases of the BP network, and a transformer fault diagnosis system is established. In addition, experiments are carried out using IEC TC 10 fault data, and the experimental results are compared with algorithms. Comparative analysis and experimental results show that the accuracy of this improved algorithm is 16%, 8%, and 24% higher than the results of Back Propagation Neural Network(BP), unimproved Manta ray foraging optimization combined with Back Propagation Neural Network(MRFO -BP), and the three-ratio method, respectively