基于神经网络的变压器故障诊断研究

摘要

Power transformer is one of the core equipments of power system, and its fault is the main factors always endangering the safety of power grids, so the study of effective fault diagnosis method has very important practical significance. A kind of neural network fault diagnosis way to transformer based on dissolved gas analysis was introduced by taking back propagation network for example. The experimental results show that the method is effective and feasible and has certain actual application value.%电力变压器是电网的核心设备之一。变压器故障一直是危及电网安全的主要因素。因此研究有效的故障诊断方法具有十分重大的现实意义。以BP网络为例。介绍了基于油中溶解气体分析的变压器神经网络故障诊断方法。试验结果表明。该方法是有效可行的，具有一定的实际应用价值。