决策树算法在装备故障检测中的应用

为快速定位装备的故障点,引入决策树方法.将故障特征作为测试属性,故障点作为类标记.采用 ID3算法计算相关信息熵对故障记录进行划分,构建所需要的决策树,用预剪枝的方法控制树的生长.以某设备故障记录样本为例,建立决策树发现故障特征和故障点之间的关联.结果表明:利用该方法对故障点进行预测,可有效缩短装备故障检测的平均时间.

Application of the Decision-Tree Algorithm to the fault diagnosis of the equipment

In order to quick location problem of the equipment, method of decision tree is introduced. The fault characteristics of failure records are treated as the tested attribute and fault point is treated as the class. It uses the ID3 decision tree algorithm of data mining theory to mining the data. Then the decision tree model, which can predict some fault point by the information of failure records, is created. The decision tree is simplified by the pre-pruning at the process of tree growth.Take some failure records of the equipment as example, the association between fault characteristics and fault point is found with decision tree algorithm. The results show that the method can predicate fault point and effectively shorten the average time for equipment fault detection.