An improved approach for fault detection by simultaneous overcoming of high-dimensionality, autocorrelation, and time-variability

Abstract

The control charts with the Principal Component Analysis (PCA) approach and its extension are among the data-driven methods for process monitoring and the detection of faults. Industrial processing data involves complexities such as high dimensionality, auto-correlation, and non-stationary which may occur simultaneously. An efficient fault detection technique is an approach that is robust against data training, sensitive to all the feasible faults of the process, and agile to the detection of the faults. To date, approaches such as the recursive PCA (RPCA) model and the moving-window PCA (MWPCA) model have been proposed when data is high-dimensional and non-stationary or dynamic PCA (DPCA) model and its extension have been suggested for autocorrelation data. But, using the techniques listed without considering all aspects of the process data increases fault detection indicators such as false alarm rate (FAR), delay time detection (DTD), and confuses the operator or causes adverse consequences. A new PCA monitoring method is proposed in this study, which can simultaneously reduce the impact of high-dimensionality, non-stationary, and autocorrelation properties. This technique utilizes DPCA property to decrease the effect of autocorrelation and adaptive behavior of MWPCA to control non-stationary characteristics. The proposed approach has been tested on the Tennessee Eastman Process (TEP). The findings suggest that the proposed approach is capable of detecting various forms of faults and comparing attempts to improve the detection of fault indicators with other approaches. The empirical application of the proposed approach has been implemented on a turbine exit temperature (TET). The results demonstrate that the proposed approach has detected a real fault successfully.

Feature selection and feature learning in machine learning applications for gas turbines: A review

“…Traditional FL adopts statistical methods mainly to reduce dimensionality and generate new features that are easier to analyse (Verleysen and François, 2005). Techniques such as PCA and kernel PCA (KPCA) have been widely and frequently applied to GT modelling problems (Hajarian et al, 2020). However, the practical challenges regarding GT modelling cannot be solved using conventional FL.…”

“…Statistical methods utilise statistical models to predict the target variables. Principal component analysis (PCA), autoregressive moving average (ARMA), and hidden Markov models are frequently applied in GT fault diagnostics and prognostics (Hajarian et al, 2020;Tahan et al, 2017). Although computationally inexpensive and capable of detecting various types of faults in engineering systems, many statistical models are based on assumptions that limit their ability to model complex behaviour (Zope et al, 2019).…”

Layered monitoring of xylenol tail gas treatment process based on stationary subspace analysis

“… is the KL divergence between and . The detailed calculation of can be found in the work of Hajarian et al [ 33 ]…”

“…[30,31] An efficient way to deal with non-stationary problems is moving window principal component analysis (MWPCA). [32][33][34] The control limits of the model are updated by removing old sample data and including new sample data, so MWPCA has a good monitoring effect on the faults occurring in non-stationary processes. With the improvement of fault detection accuracy, the monitoring of minor faults has been paid more attention.…”

“…D KL is the KL divergence between N b μ S i , c PS i and N μ S i , PS i . The detailed calculation of D KL can be found in the work of Hajarian et al[33]…”