【摘要】 针对目前研究领域中基于神经网络的变压器励磁涌流鉴别方法存在着的误判现象,提出利用高木- 关野模糊神经网络来实现鉴别的新方案。经仿真实验表明:该方案能更准确地判别出变压器内部故障和励磁涌流,具有很高的可靠性。更多还原

【Abstract】 Aiming at the wrong identification of inrush for transformer based on the artificial neural network, a new method based on the Takaji-Sugeno fuzzy neural network is presented. Distributed network, secondary harmonic, wave symmetry extent and voltage of low voltage side are used as inputs of network. The simulation tests show that this new method can correctly identify the inrush current and internal fault of transformer.更多还原