判决树方法用于变压器故障诊断的研究

提出了一种利用属于模式识别范畴的判决树ID3（Identification Data)方法进行变压器故障诊断的方法。由于ID3方法可方便地处理具有非数值特征模式且具有从样本学习规划的功能，对实例的应用显示了该手段用于变压器故障诊断的适用性。通过讲座，指出了ID3方法与人工神经网络、专家系统等人工智能方法相比所具有的优点，说明该方法和其它人工智能方法相结合是实现变压器故障诊断实用的有效途径。

ELECTRIC POWER TRANSFORMERFAULT DIAGNOSIS USING DECISION TREE

Based on ID 3 method, which belong to pattern recognition, an electric power transformer fault diagnosis method is proposed. As the ID 3 method can handle the non numerical value feature patterns and has the function that can learn decision rules from samples, the application of the proposed method to some realistic samples shows that the ID 3 method fit for transformer fault diagnosis perfectly. Through discussion, the advantages of the proposed method, compare to the other method, which use AI, such as: artificial neural nets, expert system, are pointed out. It can be seen that the combination of ID 3 method with other AI method is an effective way to the practical fault diagnosis of electric power transformer.