Real-Time Forecasting of Subsurface Inclusion Defects for Continuous Casting Slabs: A Data-Driven Comparative Study.

Subsurface inclusions are one of the most common defects that affect the inner quality of continuous casting slabs. This increases the defects in the final products and increases the complexity of the hot charge rolling process and may even cause breakout accidents. The defects are, however, hard to detect online by traditional mechanism-model-based and physics-based methods. In the present paper, a comparative study is carried out based on data-driven methods, which are only sporadically discussed in the literature. As a further contribution, a scatter-regularized kernel discriminative least squares (SR-KDLS) model and a stacked defect-related autoencoder back propagation neural network (SDAE-BPNN) model are developed to improve the forecasting performance. The scatter-regularized kernel discriminative least squares is designed as a coherent framework to directly provide forecasting information instead of low-dimensional embeddings. The stacked defect-related autoencoder back propagation neural network extracts deep defect-related features layer by layer for a higher feasibility and accuracy. The feasibility and efficiency of the data-driven methods are demonstrated through case studies based on a real-life continuous casting process, where the imbalance degree drastically vary in different categories, showing that the defects are timely (within 0.01 ms) and accurately forecasted. Moreover, experiments illustrate the merits of the developed scatter-regularized kernel discriminative least squares and stacked defect-related autoencoder back propagation neural network methods regarding the computational burden; the F1 scores of the developed methods are clearly higher than common methods.

Multi-scale Fusion of Stretched Infrared and Visible Images.

Infrared (IR) band sensors can capture digital images under challenging conditions, such as haze, smoke, and fog, while visible (VIS) band sensors seize abundant texture information. It is desired to fuse IR and VIS images to generate a more informative image. In this paper, a novel multi-scale IR and VIS images fusion algorithm is proposed to integrate information from both the images into the fused image and preserve the color of the VIS image. A content-adaptive gamma correction is first introduced to stretch the IR images by using one of the simplest edge-preserving filters, which alleviates excessive luminance shifts and color distortions in the fused images. New contrast and exposedness measures are then introduced for the stretched IR and VIS images to achieve weight matrices that are more in line with their characteristics. The IR and luminance components of the VIS image in grayscale or RGB space are fused by using the Gaussian and Laplacian pyramids. The RGB components of the VIS image are finally expanded to generate the fused image if necessary. Comparisons experimentally demonstrate the effectiveness of the proposed algorithm to 10 different state-of-the-art fusion algorithms in terms of computational cost and quality of the fused images.

Student's-t Mixture Regression-Based Robust Soft Sensor Development for Multimode Industrial Processes.

Because of multiple manufacturing phases or operating conditions, a great many industrial processes work with multiple modes. In addition, it is inevitable that some measurements of industrial variables obtained through hardware sensors are incorrectly observed, recorded or imported into databases, resulting in the dataset available for statistic analysis being contaminated by outliers. Unfortunately, these outliers are difficult to recognize and remove completely. These process characteristics and dataset imperfections impose challenges on developing high-accuracy soft sensors. To resolve this problem, the Student's-t mixture regression (SMR) is proposed to develop a robust soft sensor for multimode industrial processes. In the SMR, for each mixing component, the Student's-t distribution is used instead of the Gaussian distribution to model secondary variables, and the functional relationship between secondary and primary variables is explicitly considered. Based on the model structure of the SMR, a computationally efficient parameter-learning algorithm is also developed for SMR. Results conducted on two cases including a numerical example and a real-life industrial process demonstrate the effectiveness and feasibility of the proposed approach.

Deep Learning of Partially Labeled Data for Quality Prediction Based on Stacked Target-Related Laplacian Autoencoder.

Partially labeled data, which is common in industrial processes due to the low sampling rate of quality variables, remains an important challenge in soft sensor applications. In order to exploit the information from partially labeled data, a target-related Laplacian autoencoder (TLapAE) is proposed in this work. In TLapAE, a novel target-related Laplacian regularizer is developed, which aims to extract structure-preserving and quality-related features by preserving the feature-target mapping according to the local geometrical structure of the data. In addition, stacked TLapAE (STLapAE) is further constructed to extract deep feature representations of the data by hierarchically stacking TLapAE blocks. For model training, backward propagation equations are derived based on matrix calculus techniques to update the model parameters of the proposed TLapAE. The effectiveness of the proposed STLapAE is evaluated using the butane content prediction case in a debutanizer column, the silicon content prediction case in a blast furnace (BF) ironmaking process, and the ethane concentration prediction case in an ethylene fractionator. The results show that the proposed TLapAE model has significantly improved prediction accuracy compared to soft sensors using only labeled data and other partially labeled data modeling methods.

Deep Probabilistic Principal Component Analysis for Process Monitoring.

Probabilistic latent variable models (PLVMs), such as probabilistic principal component analysis (PPCA), are widely employed in process monitoring and fault detection of industrial processes. This article proposes a novel deep PPCA (DePPCA) model, which has the advantages of both probabilistic modeling and deep learning. The construction of DePPCA includes a greedy layer-wise pretraining phase and a unified end-to-end fine-tuning phase. The former establishes a hierarchical deep structure based on cascading multiple layers of the PPCA module to extract high-level features. The latter builds an end-to-end connection between the raw inputs and the final outputs to further improve the representation of the model to high-level features. After constructing the model structure of DePPCA, we first present the detailed training processes of the pretraining and fine-tuning stages, then clarify the theoretical merits of the proposed model from the perspective of variational inference. For process monitoring purposes, we develop two statistics based on the established DePPCA. The monitoring performance of these two statistics can remain superior even if the features extracted by DePPCA are significantly compressed to univariate. This makes the feature extraction process and online monitoring procedure of DePPCA quite fast. In other words, the proposed DePPCA can achieve accurate and efficient process monitoring by only extracting one feature for each sample. Finally, the effectiveness of DePPCA is evaluated on the Tennessee Eastman (TE) process and the multiphase flow (MPF) facility.

Security Versus Accuracy: Trade-Off Data Modeling to Safe Fault Classification Systems.

While the data-driven fault classification systems have achieved great success and been widely deployed, machine-learning-based models have recently been shown to be unsafe and vulnerable to tiny perturbations, i.e., adversarial attack. For the safety-critical industrial scenarios, the adversarial security (i.e., adversarial robustness) of the fault system should be taken into serious consideration. However, security and accuracy are intrinsically conflicting, which is a trade-off issue. In this article, we first study this new trade-off issue in the design of fault classification models and solve it from a brand new view, hyperparameter optimization (HPO). Meanwhile, to reduce the computational expense of HPO, we propose a new multiobjective (MO), multifidelity (MF) Bayesian optimization (BO) algorithm, MMTPE. The proposed algorithm is evaluated on safety-critical industrial datasets with the mainstream machine learning (ML) models. The results show that the following hold: 1) MMTPE is superior to other advanced optimization algorithms in both efficiency and performance and 2) fault classification models with optimized hyperparameters are competitive with advanced adversarially defensive methods. Moreover, insights into the model security are given, including the model intrinsic security properties and the correlations between hyperparameters and security.

Visual Process Monitoring by Data-Dependent Kernel Discriminant Analysis with t-Distributed Similarities.

Visual process monitoring would provide more directly appreciable and more easily comprehensible information about the process operating status as well as clear depictions of the occurrence path of faults; however, as a more challenging task, it has been sporadically discussed in the research literature on conventional process monitoring. In this paper, the Data-Dependent Kernel Discriminant Analysis (D2K-DA) model is proposed. A special data-dependent kernel function is constructed and learned from the measured data, so that the low-dimensional visualizations are guaranteed, combined with intraclass compactness, interclass separability, local geometry preservation, and global geometry preservation. The new optimization is innovatively designed by exploiting both discriminative information and t-distributed geometric similarities. On the construction of novel indexes for visualization, experiments of visual monitoring tasks on simulated and real-life industrial processes illustrate the merits of the proposed method.

Hessian Semisupervised Scatter Regularized Classification Model With Geometric and Discriminative Information for Nonlinear Process.

The integration of semisupervised modeling and discriminative information has been sporadically discussed in the research literature of traditional classification modeling, while the former one would make full use of the collected data and the latter one would further improve the classification performance. In this article, the Hessian semisupervised scatter regularized classification model is proposed as a coherent framework for the nonlinear process classification upon both labeled and unlabeled data. It is innovatively designed with a loss function to evaluate the classification accuracy and three regularization terms, respectively, corresponding to the geometry information, discriminative information, and model complexity. Both cases of the coherent framework, respectively, casted to the reproducing kernel Hilbert space and linear space, enjoy a theoretically guaranteed analytical solution. Experiments on process classification tasks on a benchmark dataset and a real industrial polyethylene process illustrate the merits of the proposed method in a sense that the class information of novel collected data is accurately predicted.

Deep Neural Network-Embedded Stochastic Nonlinear State-Space Models and Their Applications to Process Monitoring.

Process complexities are characterized by strong nonlinearities, dynamics, and uncertainties. Monitoring such a complex process requires a high-quality model describing the corresponding nonlinear dynamic behavior. The proposed model is constructed using deep neural networks (DNNs) to represent the state transition and observation generation, both of which constitute a stochastic nonlinear state-space model. A new bidirectional recurrent neural network (RNN), creating a connection of the hidden layer between a forward RNN and a backward RNN, is proposed to generate the filtering estimation and the smoothing estimation of process states which further generate observations with DNN-based process models. The smoothing estimator and the process model are first learned offline with all collected samples. Then the filtering estimator is fine-tuned by the learned smoother and process models to achieve real-time monitoring since the filter state is estimated based on the past and the current observations. Two indices are designed based on the learned model for monitoring the process anomaly. The proposed process monitoring model can deal with complex nonlinearities, process dynamics, and process uncertainties, all of which can be very challenging for the existing methods, such as kernel mapping and stacked auto-encoder. Two case studies validate that the effectiveness of the proposed method outperforms the other comparative methods by at least 10% when using the averaged fault detection rate in the industrial experimental data.

Semisupervised Bayesian Gaussian Mixture Models for Non-Gaussian Soft Sensor.

Soft sensors have been widely accepted for online estimating key quality-related variables in industrial processes. The Gaussian mixture models (GMM) is one of the most popular soft sensing methods for the non-Gaussian industrial processes. However, in industrial applications, the quantity of samples with known labels is usually quite limited because of the technical limitations or economical reasons. Traditional GMM-based soft sensor models solely depending on labeled samples may easily suffer from singular covariances, overfitting, and difficulties in model selection, which results in the performance deterioration. To tackle these issues, we propose a semisupervised Bayesian GMM (S2BGMM). In the S2BGMM, we first propose a semisupervised fully Bayesian model, which enables learning from both the labeled and unlabeled datasets for remedying the deficiency of infrequent labeled samples. Subsequently, a general framework of weighted variational inference is developed to train the S2BGMM, such that the rate of learning from unlabeled samples can be controlled by penalizing the unlabeled dataset. Case studies are carried out to evaluate the performance of the S2BGMM through a numerical example and two real-world industrial processes, which demonstrate the effectiveness and reliability of the proposed approach.

A neurocomputational method for fully automated 3D dendritic spine detection and segmentation of medium-sized spiny neurons.

Acquisition and quantitative analysis of high resolution images of dendritic spines are challenging tasks but are necessary for the study of animal models of neurological and psychiatric diseases. Currently available methods for automated dendritic spine detection are for the most part customized for 2D image slices, not volumetric 3D images. In this work, a fully automated method is proposed to detect and segment dendritic spines from 3D confocal microscopy images of medium-sized spiny neurons (MSNs). MSNs constitute a major neuronal population in striatum, and abnormalities in their function are associated with several neurological and psychiatric diseases. Such automated detection is critical for the development of new 3D neuronal assays which can be used for the screening of drugs and the studies of their therapeutic effects. The proposed method utilizes a generalized gradient vector flow (GGVF) with a new smoothing constraint and then detects feature points near the central regions of dendrites and spines. Then, the central regions are refined and separated based on eigen-analysis and multiple shape measurements. Finally, the spines are segmented in 3D space using the fast marching algorithm, taking the detected central regions of spines as initial points. The proposed method is compared with three popular existing methods for centerline extraction and also with manual results for dendritic spine detection in 3D space. The experimental results and comparisons show that the proposed method is able to automatically and accurately detect, segment, and quantitate dendritic spines in 3D images of MSNs.

Ensemble semi-supervised Fisher discriminant analysis model for fault classification in industrial processes.

In this paper, an ensemble form of the semi-supervised Fisher Discriminant Analysis (FDA) model is developed for fault classification in industrial processes. This method uses the K Nearest Neighbor (KNN) algorithm to merge the metric level outputs, which are obtained by the sub-classifiers in the ensemble model, to get the final classification result. An adaptive form is further proposed to improve the classification performance by putting forward to a new method of weight adjustment. While semi-supervised learning can generate a better model by exploiting additional information contained in unlabeled data, ensemble learning achieves the promotion of algorithm robustness by integrating a series of weak learners. In addition, the property of diversity in ensemble learning can be boosted by incorporating different unlabeled data to different weak learners. Therefore, the combination of those two methods can achieve great generalization for the fault classification model. The performances of two proposed methods are evaluated through an industrial benchmark process.

[The online measurement for pulp Kappa number based on near infrared spectroscopy and support vector machine].

A new method for online measurement of pulp Kappa number by means of near infrared diffuse reflectance spectroscopy and support vector machine (SVM) modeling has been developed in this paper. The near infrared diffuse reflectance spectroscopy of 45 Chinese red pine wood pulp samples was acquired. Selecting the absorption rates in 15 vibration absorption peaks of each sample and using dynamic independent component analysis (DICA) to distill the characters of input sample data, the pulp Kappa number predictive model based on SVM was built. From the whole 45 samples, 35 samples was selected to be the calibration set, and the predictive set consisted of the other 10 samples was used to validate the the pulp Kappa number predictive model. The external validation standard deviation is 0.26 for pulp Kappa number predictive model based on SVM, and the determining factor is 0.93 for the model. The internal cross validation standard deviation is 0.22 for pulp Kappa number predictive model based on SVM, and the determining factor is 0.96 for the model. To analyze the effectiveness of SVM method used to build the pulp Kappa number predictive model, the pulp Kappa number predictive model based on linear regression (LR) was also established. The external validation standard deviation is 0.45 for the model based on linear regression (LR), and the determining factor is 0.81 for the model. The internal cross validation standard deviation is 0.41 for the model based on linear regression (LR), and the determining factor is 0.85 for the model. For the 10 test samples, the pulp Kappa number predictive model based on Linear regression (LR) and the model based on SVM all have certain predictive accuracy, but the later higher. The experiment results not only show the feasibility and effectiveness of the near infrared measurement method for pulp Kappa number, but also validate that the pulp Kappa number predictive model based on SVM is more accurate and robust than linear regression model.

Active surge control for variable speed axial compressors.

This paper discusses active surge control in variable speed axial compressors. A compression system equipped with a variable area throttle is investigated. Based on a given compressor model, a fuzzy logic controller is designed for surge control and a proportional speed controller is used for speed control. The fuzzy controller uses measurements of the change of pressure rise as well as the change of mass flow to determine the throttle opening. The presented approach does not require the knowledge of system equilibrium or the surge line. Numerical simulations show promising results. The proposed fuzzy logic controller performs better than a backstepping controller and is capable to suppress surge at different operating points.

Coronary artery segmentation using geometric moments based tracking and snake-driven refinement.

Automatic or semi-automatic segmentation and tracking of artery trees from computed tomography angiography (CTA) is an important step to improve the diagnosis and treatment of artery diseases, but it still remains a significant challenging problem. In this paper, we present an artery extraction method to address the challenge. The proposed method consists of two steps: (1) a geometric moments based tracking to secure a rough centerline, and (2) a fully automatic generalized cylinder structure-based snake method to refine the centerlines and estimate the radii of the arteries. In this method, a new line direction based on first and second order geometric moments is adopted while both gradient and intensity information are used in the snake model to improve the accuracy. The approach has been evaluated on synthetic images as well as 8 clinical coronary CTA images with 32 coronary arteries. Our method achieves 94.7% overlap tracking ability within an average distance inside the vessel of 0.36 mm.

Multi scale and slice-based approach for automatic spine detection.

Dendritic spines play an essential role in the central nervous system. Recent experiments have revealed that neuron functional properties are highly correlated with the statistical and morphological changes of the dendritic spines. In this paper, we propose a new multi scale approach for detecting dendritic spines in a 2D Maximum Intensity Projection (MIP) image of the 3D neuron data stacks collected from a 2-photon laser scanning confocal microscope. The proposed method utilizes the curvilinear structure detector in conjunction with the multi scale spine detection algorithm which automatically and accurately extracts and segments the spines with variational sizes along the dendrite. In addition, a slice-based spine detection algorithm is also proposed to detect spines which are hidden from the MIP image within the dendrite area. Experimental results show that our proposed method is effective for automatic spine detection and is able to accurately segment dendrite.