Fault diagnosis of transformers based on complementary immune algorithm

作者：Yuan Jinsha; Zhang Liwei^*; Li Zhong; Zhang Yinghui

来源：Transactions of China Electrotechnical Society, 2015, 30(24): 67-75.

• Artificial Immune System
• Complementary strategy
• Data characteristics
• Fault diagnosis of transformers
• Immune operator
• Intelligence algorithms
• Particle swarm optimization algorithm
• Transformer fault diagnosis

The transformer fault diagnosis based on self-organization antibody net (soAbNet) has no network compression mechanism and selects the initial antibodies randomly, so its network performance is instable. Thus, a diagnosis method based on complementary immune algorithm for power transformer is proposed in this paper, and immune operator is designed in detail considering the characteristics of transformer fault diagnosis. Vaccination of immune operator uses K-means optimal clustering algorithm to provide initial antibodies for soAbNet and compresses the network through immune selection, and its parameter is optimized by particle swarm optimization (PSO) algorithm. Experimental results demonstrate that the proposed complementary immune algorithm could make use of prior knowledge and extract the data characteristics of the fault samples effectively. The diagnostic accuracy of the proposed algorithm is higher than that of the single intelligence algorithm.

• 出版日期2015
• 单位华北电力大学