基于PCA的氧化铝蒸发系统故障检测  

Application of PCA-Based Fault Detection in Alumina Evaporator System

作　　者：胡志坤[1] 李哲彬 陈志文[2] 

机构地区：[1]中南大学物理与电子学院,长沙410083 [2]杜伊斯堡-埃森大学自动化控制与复杂系统学院

出　　处：《上海应用技术学院学报（自然科学版）》2015年第3期221-226,共6页Journal of Shanghai Institute of Technology: Natural Science

摘　　要：针对氧化铝蒸发系统结构复杂、物理模型难以搭建、大量数据得不到合理利用的问题,提出了基于主成分分析(PCA)的故障检测方法和直观的故障分离方法.通过对氧化铝蒸发系统进行深入分析,对系统故障进行分类,并在仿真模型中建立了不同的故障类型模型.最后,基于对氧化铝蒸发故障进行模拟得到的故障数据,给出了在氧化铝蒸发过程故障检测中的实例,验证了PCA方法的可行性.In view of the problem of complex structure and difficulty in building the physical model and the failure of rational utilization of large amount of data for the alumina evaporation system, a fault detection method and a fault isolation method based on principal component analysis （PCA） were proposed. Through analysing the alumina evaporation system thoroughly, fault types were classified and fault models were carried out. Finally, based on the simulated fault data obtained for alumina evaporation model, an example was presented in the fault detection of alumina evaporation process to verify the feasibility of PCA method.

关 键 词：氧化铝蒸发系统 主成分分析(PCA) 故障检测 故障分离 

分 类 号：TQ133.1[化学工程—无机化工]