基于灰度聚类的钢管水压试验机故障诊断  

Fault Diagnosis Based on Grey Clustering Algorithm for Hydraulic Tube Tester

作　　者：胡学发[1] 赵珍[1] 王福利[1] 何大阔[1] 

机构地区：[1]东北大学流程工业综合自动化教育部重点实验室,辽宁沈阳110004

出　　处：《东北大学学报（自然科学版）》2007年第11期1529-1532,共4页Journal of Northeastern University(Natural Science)

基　　金：国家自然科学基金资助项目(60374003);国家重点基础研究发展规划项目(2002CB312201)

摘　　要：针对钢管水压试验机在水压试验过程中可能出现各类复杂故障,并且短时间内很难找到故障发生的地点和原因的现象,提出了一种改进的灰度聚类方法,该方法通过提取反映真实打压过程的数据对生产过程中的打压曲线进行聚类分析,并对聚类结果应用专家系统进行推理分析,找出相应的故障结论.为了验证该方法的有效性和可靠性,将该方法在某钢厂3号线水压试验机上实现,实际结果表明该方法可靠有效地实现了对各种故障的划分和诊断.Pressure testing in hydraulic tube tester may induce a wide range of possible failures to reduce productivity, and it is hard to find locations and reasons when they happen. An improved grey clustering algorithm is therefore presented to cluster the data of hydraulic tube tester by selecting the data reflecting real production, then the expert system is used to diagnose the results of this improved grey clustering algorithm. This approach is applied to the #3 hydraulic tube tester in a corporation. The experiment results demonstrated that the accuracy can be improved effectively through this new approach.

关 键 词：钢管 水压试验机 故障诊断 灰度聚类算法 专家系统 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]