基于数据挖掘的空调机组故障预测研究  

Study on Error Prediction in Air Conditioner Set Based on Data Mining

作　　者：张积洪[1] 王慧敏[1] 陈维兴[1] 

机构地区：[1]中国民航大学航空自动化学院,天津300300

出　　处：《测控技术》2015年第8期37-40,44,共5页Measurement & Control Technology

基　　金：中央高校基本科研业务中国民航大学专项(3122013C015);中央高校基本科研业务民航节能减排专项(ZXH2012G005)

摘　　要：作为机场桥载设备的重要组成部件之一——空调机组,由它产生的频发故障所造成的设备利用率低、修复率时长和重大经济损失等问题,在基于桥载设备的安监系统上,设计了空调机组安监信息采集节点,通过Clementine软件,建立了空调机组的故障预测模型,经过比较Apriori算法和其他典型数据挖掘算法的性能,通过在线数据库测试结果表明,得到了不同的典型算法在预测空调机组故障上显现的特点,通过选择最优算法高效地预测了空调机组的未来状态,实现了对空调机组的实时故障预测,进而为解决故障提供了方向和目标,最终达到了降低经济损失最大化的目的,具有很深的实际意义。Air conditioner is one of the most important part of the bridge-born equipment on airport.By structuring the model of error prediction and the collecting node of safety monitoring information with the Clementine software,frequent errors appeared in air conditioner can be effectively avoided.The frequent errors would usually give rise to a series of problems,including low utilization rate,long repair rate time and severely economic losses and so on.By comparing performance of Apriori algorithm with other typical algorithms,on the basis of online database,the obvious characteristics emerged in predicting the future state of air conditioner are obtained.Moreover,the future state of air conditioner is effectively predicted by selecting the optimal algorithm to achieve the real-time failure prediction.Furthermore,the model can provide failure solution with direction and goal.Finally,the purpose that economic loss can be reduced to maximization is gotten,which will have considerably practical significance.

关 键 词：桥载设备 安监系统 频繁项集 支持度 置信度 

分 类 号：TP319[自动化与计算机技术—计算机软件与理论]