基于 ANN 状态识别系统的变压器故障检测方法

A Detecting Method of Winding Faults Based on ANN Type State Recognition System for Transformer Protection

提出基于ＡＮＮ状态识别系统的变压器故障检测方法。该系统主要由ＡＮＮ变压器一次电压非线性估计器、比较器、计数器等构成。它通过比较一次电压实际值与估计值的大小，识别变压器内部是否发生故障。关于ＡＮＮ状态识别系统在检测变压器故障中的正确性及快速性，通过数字仿真得到验证。

The paper presents a new fault recognition method for transformer protection on artificial network(ANN) It is comprised of a nonlinear estimator,which is based in ANN approach and used to estimate the primary voltage of a transformer,a comparator and a counter The internal fault of a transformer is checked by comparing estimated and actual value of primary voltage of the transformer The digital simulation result proves that the method is correct and fast