摘要
针对复杂设备故障预警弊端,提出一种含有隐性故障的复杂设备多源关联故障预警方法;因为复杂设备监控点较多,监控数据通常以大量、快速、时变的流形式持续到达,首先构建了大数据智能平台将监控数据分流为在线处理和离线处理;其次在研究隐性故障和显性故障的关联性、因果性的基础上,提出了多源关联自适应故障预警模型;最后提出了启发式故障频繁挖掘算法,对链接条件、剪枝条件进行了约束;实验表明该方法在预警准确率上明显优于传统预警方法,且随着故障数的增加,预警准确率较为稳定,适用于含有隐性故障的大型复杂设备的故障预警与诊断。
Multi--source correlation failure warning method of the complex equipment with hidden failure has been proposed in order to avoid the drawbacks of failure warning of the complex equipment with bidden failure. The monitoring data usually continue to arrive in mass, rapid and time--varying way. So, the intelligent platform of big data has been built to handle the monitoring data by divide them into online part and off--line part firstly. Secondly, Multi--source Correlation Adaptive Fault Warning Model has been proposed based on the research on the correlation and causality of hidden failure and overt failure. Lastly, the link condition and pruning condition are further con- strained by Heuristic Failure Frequent Mining Algorithm. Experiment showed that Multi--source correlation failure warning method of the complex equipment with hidden failure is obviously better than the traditional warning method. By the number of fault increased, the accu- racy rate of failure warning is relatively stable. It is appropriate for failure warning and diagnosis of the complex equipment with hidden failure
出处
《计算机测量与控制》
北大核心
2014年第4期1030-1032,共3页
Computer Measurement &Control
基金
河北省自然科学基金项目(G2010001331)
河北省自然科学基金项目(G2014202031)
河北省科技厅软科学项目(13451813D)
河北省高等学校科学技术研究项目(QN20131060)
关键词
复杂设备
隐含故障
故障诊断
故障预警
关联规则算法
complex equipment
hidden failure
failure diagnosis
failure warning
association rules arithmetic