阴性选择算法在工业控制系统故障诊断中的应用  

Application of negative selection algorithm in fault diagnosis for industrial control system

作　　者：牛慧峰[1] 姜万录[1] 王文杰 

机构地区：[1]燕山大学重型机械流体动力传输与控制河北省重点实验室,河北秦皇岛066004 [2]大庆艺朗司特石油专用管有限责任公司,黑龙江大庆163100

出　　处：《燕山大学学报》2008年第4期362-366,共5页Journal of Yanshan University

基　　金：国家自然科学基金项目(50775198);河北省教育厅博士基金资助项目(B2004128)

摘　　要：受生物免疫系统的启发,把免疫原理应用到故障诊断领域。分析了现有阴性选择算法在故障诊断应用中的不足之处,提出了改进后的阴性选择算法。介绍了田纳西-伊斯曼(TEP)过程仿真系统,用改进后的阴性选择算法对该系统的21种故障样本进行仿真试验,各故障的平均检出率达到85%以上,验证了此算法在工业控制系统故障诊断中的有效性。Inspired by artificial immune system, the immune principle is applied to the fault diagnosis field. The improved negative selection algorithm was expatiated, after analyzing the shortcoming of the existent negative selection algorithm. The simulated system for Tennessee Eastman Process was introduced and the 21 fault samples of this system were diagnosed using the improved negative selection algorithm. The correct rate of the diagnosis is above 85%. It is proved that the improved algorithm is valid to diagnose faults of industry control system.

关 键 词：阴性选择算法 故障诊断 田纳西-伊斯曼过程 人工免疫系统 

分 类 号：TP306[自动化与计算机技术—计算机系统结构]