Transition Process Modeling and Monitoring Based on Dynamic Ensemble Clustering and Multiclass Support Vector Data Description

Data-driven approaches are widely applied for fault detection in industrial process. Recently, a new method for fault detection called principal component pursuit(PCP) is introduced. PCP is not only robust to outliers, but also can accomplish the objectives of model building, fault detection, fault isolation and process reconstruction simultaneously. PCP divides the data matrix into two parts: a fault-free low rank matrix and a sparse matrix with sensor noise and process fault. The statistics presented in this paper fully utilize the information in data matrix. Since the low rank matrix in PCP is similar to principal components matrix in PCA, a T2 statistic is proposed for fault detection in low rank matrix. And this statistic can illustrate that PCP is more sensitive to small variations in variables than PCA. In addition, in sparse matrix, a new monitored statistic performing the online fault detection with PCP-based method is introduced. This statistic uses the mean and the correlation coefficient of variables. Monte Carlo simulation and Tennessee Eastman (TE) benchmark process are provided to illustrate the effectiveness of monitored statistics.

Multimode processes may be operating under different statuses, revealing different process characteristics and variations. It can be an interesting issue to check how the process variations change from one mode to another and thus how they can be used for multimode analysis and monitoring. In the present work, a between-mode relative analysis algorithm is proposed to explore the between-mode relationship. Comprehensive subspace decomposition is performed in each mode regarding their relative changes from the reference mode and influences on process monitoring. Each mode is separated into three different systematic subspaces and one residual subspace based on the between-mode changes of process variations, including the increased part, the decreased part, and the unchanged part. These different variations are then modeled respectively and different monitoring statistics are defined for online monitoring to identify mode affiliation and detect the fault status. By separating and monitoring different types of between-mode relative variations, the proposed method can not only efficiently distinguish different modes and detect faults but also offer enhanced process understanding. Theoretical support is framed and the related statistical characteristics are analyzed. Its feasibility and performance are illustrated with multimode data generated from the Tennessee Eastman (TE) benchmark process.

Fuzzy decision fusion system for fault classification with analytic hierarchy process approach

Sensor faults are non-negligible issues for soft sensor modeling. However, existing deep learning-based soft sensors are fragile and sensitive when considering sensor faults. To improve the robustness against sensor faults, this article proposes a deep subdomain learning adaptation network (DSLAN) to develop a sensor fault-tolerant soft sensor, which is capable of handling both sensor degradation and sensor failure simultaneously. Primarily, domain adaptation works for process data with sensor degradation in industrial processes. Being founded on the basic structure of deep domain adaptation, a novel subdomain learner is added to automatically learn the subdomain division, enabling DSLAN adaptable to multimode industrial processes. Notably, the subdomain structure of each sample follows a categorical distribution parameterized by output of the subdomain learner. Based on the designed subdomain learner, a new probabilistic local maximum mean discrepancy (PLMMD) is presented to measure the difference in distribution between source and target features. In addition, a generator for failure data imputation is integrated in the framework, making DSLAN handle sensor failure simultaneously. Finally, the Tennessee Eastman (TE) benchmark process and two real industrial processes are used to verify the effectiveness of the proposed method. With the fault tolerance ability, soft sensing technology will take a step toward practical applications.

Switching LDS-based approach for process fault detection and classification

Nonlinear process monitoring based on linear subspace and Bayesian inference

Usually, industrial processes have multiple operational modes due to different production strategies, external environmental variability, or changes in product specifications. Monitoring of multimode processes constitutes a challenging problem considering multiple steady-state operational regions and dynamic transitions. This paper proposes a novel method for the offline identification of stable modes and transitions based on a local kernel density estimation algorithm. The online monitoring scheme is based on mode identification and transition tracking. The Tennessee Eastman (TE) benchmark process is used as a case study to evaluate the performance of the proposal. As a result, stable modes are successfully isolated from transitions, even when these involve complex changes in the production mode. The results also demonstrate that the proposed scheme is capable of tracking mode changes, and finally, results monitored during transitions confirm the validity and efficacy of the new approach compared with pr...

In the majority of the multivariable processes, analysis of process monitoring and fault diagnosis is usually based on the fundamental assumption that the monitored variables follow a Gaussian distribution. However, it is well known that many of the variables are mutually dependent in process systems. This paper proposes a new monitoring method based on independent component analysis (ICA) - sparse autoencoder. The independent information component can be extracted by ICA through higher-order statistics. Moreover, the inherent nonlinear characteristics in the residual model of ICA can be handled by a deep architecture constructed with sparse autoencoder. To overcome the problem of local minima in the optimization of sparse autoencoder, a restricted Boltzmann machine (RBM) is used to pre-train the net, and the parameters in the sparse autoencoder is updated by Limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS). Monitoring statistic is developed and its confidence limit is computed by kernel density estimation. A case study of the Tennessee Eastman (TE) benchmark process indicates that the proposed fault detection method is more efficient.

The problem of fault classification in industry has been studied extensively. Most classification algorithms are modeled on the premise of data balance. However, the difficulty of collecting industrial data in different modes is quite different. This inevitably leads to data imbalance, which will adversely affect the fault classification performance. This article proposes a novel data augmentation classifier (DAC) for imbalanced fault classification. Data augmentation based on generative adversarial networks (GANs) is an effective way to solve the problem of unbalanced classification. However, the randomness of the GAN generation process restricts the effect of data enhancement. DAC proposes a data selection strategy based on data filtering and data purification in model training to solve this problem. In addition, DAC combines supervised learning and data generation processes to obtain an end-to-end model. Meanwhile, multigenerator structure of DAC (MDAC) is proposed to solve the problem of incomplete learning of a single generator when data imbalances get complicated. The proposed DAC and MDAC are applied in two fault classification cases of the Tennessee Eastman (TE) benchmark process, results of which show superiority of DAC and MDAC compared to existing methods. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —Data imbalances are common in fault classification and affect the effectiveness of modeling in industry. As a generative model, generative adversarial networks (GANs) provide new ideas for small-class data augmentation. However, the instability of its training process and the randomness of data generation affect the results of data augmentation. In this article, the GAN generation process is analyzed in detail. The results of the visualization indicate that no data generation was perfect at any one time. Based on the rules of GAN data generation, we propose a data selection strategy during training. High-quality data are selected for data augmentation through data filtering and data purification. Apart from this, we combine the training process of GAN and classification model for imbalanced data to reduce modeling time. Through industrial examples, we have evaluated the effectiveness of this method.

Fault detection in dynamic plant-wide process by multi-block slow feature analysis and support vector data description

Quality-related fault detection for dynamic process based on quality-driven long short-term memory network and autoencoder

In recent years, partial least squares (PLS) has become more and more closely linked with the monitoring of key performance indicators (KPI) in large-scale industrial processes. However, the standard PLS and its modifications still have some problems during process monitoring, because these kinds of methods cannot fully utilize the correlation between variables. Instead of using the traditional KPI-related PLS, this paper uses canonical correlation analysis (CCA) to exploit the best correlation among variables. The proposed method decomposes process data into KPI-related and KPI-unrelated parts, and then constructs statistics for process monitoring. For short, the proposed method is termed as KPI-CCA. Finally, the effectiveness and advantages of KPI-CCA are validated by a numerical example and the Tennessee Eastman (TE) benchmark process simulation.

In this paper, an indirect optimization criterion for parameter setting the kernel-based fault detection process is applied. The procedure analyzed involves the data preprocessing through the Kernel Independent Component Analysis (KICA) method, and the fault detection by using a classifier based on the Kernel Fuzzy C-means (KFCM) algorithm to reduce the classification errors. The main objective of the paper is the adjustment of the kernel parameters to obtain the best possible performance in the fault detection. To achieve this, two different metaheuristic algorithms are used: Differential Evolution and Particle Swarm Optimization. The proposed approach was evaluated by using the Tennessee Eastman (TE) process benchmark.

Kernel principal component analysis (KPCA) has become a popular technique for process monitoring in recent years. However, the performance largely depends on kernel function, yet methods to choose an appropriate kernel function among infinite ones have only been sporadically touched in the research literatures. In this paper, a novel methodology to learn a data-dependent kernel function automatically from specific input data is proposed and the improved kernel principal component analysis is obtained through using the data-dependent kernel function in traditional KPCA. The learning procedure includes two parts: learning a kernel matrix and approximating a kernel function. The kernel matrix is learned via a manifold learning method named maximum variance unfolding (MVU) which considers underlying manifold structure to ensure that principal components are linear in kernel space. Then, a kernel function is approximated via generalized Nystrom formula. The effectiveness of the improved KPCA model is confirmed by a numerical simulation and the Tennessee Eastman (TE) process benchmark.

Process monitoring via enhanced neighborhood preserving embedding