决策树分类算法在发电设备状态检修中的应用研究  

APPLICAION OF DECISION TREE CLASSIFICATION ALGORITHM IN CONDITION-BASED MAINTENANCE OF POWER GENERATION EQUIPMENTS

作　　者：李凡生[1] 陈庆吉[1] 

机构地区：[1]东北电力学院管理系,吉林省吉林市132012

出　　处：《电网技术》2003年第12期67-70,共4页Power System Technology

摘　　要：针对当前火力发电厂实施设备状态检修面临的问题,提出了用数据挖掘中的决策树分类算法对反映设备状态的各类指标数据进行分析,挖掘设备状态的分类规则。对如何构造设备状态训练样本集提出了可行办法,详细描述了运用SQL语言实现决策树分类算法(ID3)的具体步骤。通过实例计算验证了该算法的正确性和有效性。According to the problems to be faced with the implementation of condition-based maintenance of power generation equipments it is pointed out to analyze the various kinds of data reflecting the equipment conditions by the decision tree classification algorithm in the data mining and to mine the classification rules of equipment conditions. A feasible method, i.e., how to construct the training sample set of equipment conditions, is put forward. The concrete procedures to implement the decision tree algorithm by SQL language are described in detail. By means of practical examples the correctness and effectiveness of the presented algorithm are proved.

关 键 词：火力发电厂 发电设备 状态检修 决策树分类算法 

分 类 号：TM621.3[电气工程—电力系统及自动化]