基于Fuzzy ART网络识别变压器绕组过热故障

Recognition on fault of transformer coil hot-spot using Fuzzy ART

提出了基于模糊自适应共振理论神经网络识别变压器绕组热点故障的新方法。该方法可以识别随机模糊模式 ,尤其适用于及时捕捉变压器负荷突变、油中特征气体含量突发性异常等偶发性故障征兆 ,对及时排除绕组热点的偶发故障隐患具有重要意义。实例分析表明 。

A novel methodology to identify the hot\|spot of transformer coil based on Fuzzy ART(adaptive resonance theory)neural network is suggested.A random fuzzy pattern can be identified using the methodology.It is suitable especially to catch some random faults,such as sudden changing load current and sudden changing the composition of the characteristic gases dissolved in transformer oil.It is important to eliminate the random fault on the hot\|spot of transformer coil.Experimental results indicate this methodology is effective.