基于自适应控制限的纵向标准化主元分析故障检测策略  

Longitudinal-standardization PCA fault detection strategy based on adaptive confidence limit

作　　者：王天真[1] 倪孟骐 吴昊[1] 汤天浩[1] 

机构地区：[1]上海海事大学物流工程学院,上海201306

出　　处：《上海海事大学学报》2016年第1期92-96,共5页Journal of Shanghai Maritime University

基　　金：国家自然科学基金(61203089;61304186;61403229);国家高技术研究发展计划("八六三"计划)(2013A2041106);上海市教育委员会重点创新项目(14ZZ141)

摘　　要：为有效地解决在非稳定工况时信号具有非高斯分布及突变特点的故障检测问题,针对周期非稳定工况提出一种基于自适应控制限的纵向标准化主元分析故障检测策略.该策略主要包含两部分:一是将非高斯数据转换成高斯数据;二是构建自适应控制限,解决信号突变问题.仿真结果验证了该策略对周期非稳定工况故障检测的可行性和有效性.To efficiently solve fault detection problems that signals show characteristics of non-Gaussian distribution and abrupt change in unsteady-stage conditions,a longitudinal-standardization Principal Component Analysis（ PCA） fault detection strategy based on adaptive confidence limit is proposed for periodic unsteady-stage conditions. The strategy comprises two parts： the first part is to convert the non-Gaussian data to Gaussian data; the second one is to build the adaptive confidence limit to solve abrupt change problems of signals. The proposed strategy is feasible and effective for fault detection in periodic unsteady-stage conditions,which is validated by simulation results.

关 键 词：故障检测 主元分析(PCA) 纵向标准化 自适应控制限 

分 类 号：TP277.3[自动化与计算机技术—检测技术与自动化装置]