摘要
针对工业过程的故障诊断问题,文中提出了一种基于小波去噪结合规范变量分析与独立元分析(CVA-ICA)的故障检测方法。该方法首先利用小波变换对过程数据去噪,然后用CVA方法求出观测数据的规范变量,并且对规范变量进行ICA分解,最后运用和统计量判断工况是否正常。通过对Tennessee Eastman(TE)过程的仿真研究,验证了该方法的可行性与有效性。
In order to handle the problem of fault diagnosis for industrial processes, an improved fault detection method was proposed based on wavelet de-noise integrated with canonical variable analysis （CVA）and independent component analysis （ICA）. The wavelet transform was used for de-noising,and the CVA algorithm wass used to calculate the canonical variable of the data, and then, the ICA algorithm was used to decompose the canonical variable. The fault detection was performed by means of Ho- tellingand. A case study of Tennessee Eastman （TE） process shows that the proposed algorithm is feasible and efficient.
出处
《仪表技术与传感器》
CSCD
北大核心
2014年第4期80-84,共5页
Instrument Technique and Sensor
关键词
小波去噪
规范变量分析
独立元分析
故障检测
wavelet de-noise
canonical variable analysis
independent component analysis
fault detection
