Journal of Computer Applications ›› 2016, Vol. 36 ›› Issue (12): 3505-3510.DOI: 10.11772/j.issn.1001-9081.2016.12.3505

Previous Articles     Next Articles

Prediction for intermittent faults of ground air conditioning based on improved Apriori algorithm

CHEN Weixing, QU Rui, SUN Yigang   

  1. College of Electronic Information and Automation, Civil Aviation University of China, Tianjin 300300, China
  • Received:2016-04-13 Revised:2016-06-30 Online:2016-12-10 Published:2016-12-08
  • Supported by:
    This work is partially supported by the Civil Aviation Joint Research Funds of National Science Council and China Civil Aviation Administration (U1433107), the Fundamental Research Funds for the Central Universities (3122014D021).


陈维兴, 曲睿, 孙毅刚   

  1. 中国民航大学 电子信息与自动化学院, 天津 300300
  • 通讯作者: 曲睿
  • 作者简介:陈维兴(1981-),男,天津人,副教授,硕士,主要研究方向:嵌入式系统、工业系统网络;曲睿(1990-),女,辽宁沈阳人,硕士研究生,主要研究方向:数据融合、数据挖掘;孙毅刚(1963-),男,山东汶上人,教授,博士,主要研究方向:航空设备智能监控。
  • 基金资助:

Abstract: Aiming at the problems caused by intermittent faults of ground air conditioning, including low use efficiency, maintenance lag etc., a prediction method of intermittent faults which combined re-association Array Summation (AS)-Apriori with clustering K-means was raised, based on this method, delayed maintenance forecast was realized. The low efficiency problem of frequently scanning transaction database in Apriori was solved in AS-Apriori algorithm, by constructing intermittent fault arrays and giving a summation of corresponding items on them in real-time. The goal of delayed maintenance forecast is to estimate the critical region of permanent fault to arrange reasonable maintenance, which can be realized by using Gaussian distribution for the solution of maintenance wave of different intermittent fault variables and delay probability and then giving an accumulation in order. The results show that, the operational efficiency is improved, the support degree of re-association rules is upgraded by 20.656 percentage points, and more accurate prediction of intermittent failure is realized. Moreover, according to the analysis of data, the probability of forecasting maintenance-wave and delay-probability is shown as a linear distribution, which means that the high predictability of intermittent faults is more convenient to maintain and manage in advance and the formation of permanent fault is reduced.

Key words: ground air conditioning, data mining, intermittent fault, prediction, permanent fault

摘要: 针对机坪地面空调间歇故障引起的使用效能低、维修滞后等问题,提出了二次关联累加数组(AS)-Apriori与聚类K-means相结合的间歇故障预测方法,并基于此实现了延误维修预测。其中:AS-Apriori算法解决了Apriori频繁扫描事务库的低效问题,通过实时构造间歇故障数组并对其对应项累加求和;延误维修预测是为了估计出永久故障临界区以安排合理维修,可采用正态分布求出不同间歇故障变量的维修波及延误概率并进行依次累加而实现。验证表明,AS-Apriori提高了运行效率,且二次关联规则支持度提升了20.656个百分点,能更准确预测间歇故障,同时参照数据分析,预测的维修波及延误累加概率呈线性分布,即可预测性高的间歇故障更便于预先维护管理,减少永久故障的形成。

关键词: 地面空调, 数据挖掘, 间歇故障, 预测, 永久故障

CLC Number: