Canonical Variate Dissimilarity Analysis for Process Incipient Fault Detection

Abstract

Early detection of incipient faults in industrial processes is increasingly becoming important, as these faults can slowly develop into serious abnormal events, an emergency situation, or even failure of critical equipment. Multivariate statistical process monitoring methods are currently established for abrupt fault detection. Among these, canonical variate analysis (CVA) was proven to be effective for dynamic process monitoring. However, the traditional CVA indices may not be sensitive enough for incipient faults. In this work, an extension of CVA, called the canonical variate dissimilarity analysis (CVDA), is proposed for process incipient fault detection in nonlinear dynamic processes under varying operating conditions. To handle non-Gaussian distributed data, kernel density estimation was used for computing detection limits. A CVA dissimilarity-based index has been demonstrated to outperform traditional CVA indices and other dissimilarity-based indices, namely DISSIM, RDTCSA, and GCCA, in terms of sensitivity when tested on slowly developing multiplicative and additive faults in a CSTR under closed-loop control and varying operating conditions.

Mixed kernel canonical variate dissimilarity analysis for incipient fault monitoring in nonlinear dynamic processes

“…CVDA is a framework based on canonical variate analysis (CVA), which is an effective dynamic MSPM method (Odiowei and Cao, 2010). CVDA aims to enhance CVA for incipient fault detection (Pilario and Cao, 2018). In this paper, the proposed MK-CVDA consists of KPCA followed by CVDA.…”

“…Cheng et al (2010) and Ji et al (2018) used the multivariate exponentially weighted moving average for capturing small mean shifts in the process. Recently, Pilario and Cao (2018) proposed Canonical Variate Dissimilarity Analysis (CVDA) to detect incipient faults even at dynamically varying process operating conditions. However, the nonlinear issue needs to be handled more efficiently in CVDA, since processes are inherently nonlinear in practice.…”

A Review of Kernel Methods for Feature Extraction in Nonlinear Process Monitoring

“…Nevertheless, many researchers have derived analytical expressions for either kernel contributions-based diagnosis [66,79,81,83,87,94,119,127,133,136,146,150,156,157,162,164,194,213,241,268,275,276,278,279,288,289,293] or kernel reconstructions-based diagnosis [86,117,140,155,161,163,176,217,236,254,265,285]. However, most derivations are applicable only when the kernel function is the RBF, Equation (5). In one approach, Tan and Cao [251] proposed a new deviation contribution plot to perform fault identification for any nonlinear feature extractor.…”

“…Fault detection, diagnosis, and prognosis methods aim to, respectively, determine the presence, identify the cause, and predict the future behavior of these process anomalies [2,4]. Thus, process monitoring is a key layer of safety for maintaining an efficient and reliable operation of industrial plants [5].…”

Large‐scale dynamic process monitoring based on performance‐driven distributed canonical variate analysis

“…As one of the subspace techniques, canonical variate analysis (CVA) is very suitable to model the dynamic process, since it not only utilizes the state-space representation to capture the dynamic information in a very accurate way, but also provides optimal predictors between the future and the past information. Recently, various CVA-based models have been developed and the effectiveness of CVA has been adequately verified [17][18][19] . Thus, CVA is also introduced into our work to model the dynamic information for each subblock.As to the third problem, on the one hand, for the multiblock modeling of the large-scale process, if a fault is easily to be detected, it probably occurs that the majority or even all of the subblocks are affected by the fault.…”