In the blast furnace iron-making process (BFIP), there still has been a significant push to maintain a stable process and ensure maximum efficiency. Although some control systems can compensate for multiple types of disturbances, some significant process faults always require precise human intervention to avoid safety hazards. Therefore, it is crucial to develop an efficient and stable diagnosis system to identify these faults quickly. This article focuses on a novel approach called deep stationary kernel learning support vector machine (DSKL-SVM) for BFIP fault diagnosis. To eliminate the impact of nonstationary property on modeling, analytic stationary subspace analysis (ASSA) is adopted to estimate consistent underlying features. Then, design a multilayer stacked deep kernel network to explore deep nonlinear information further. A DSKL-SVM classifier and the corresponding two-tier model optimization algorithm are constructed to isolate different data types. At last, a series of case studies based on actual BFIP present the effectiveness of DSKL-SVM. The results demonstrate that the proposed method has an outstanding effect on fault diagnosis, and it is verified that the performances of stationary projection (SP), deep stationary nonlinear (DSN) construction, and online computation times are superior to other methods.

For the actual blast furnace ironmaking process (BFIP), sophisticated dynamic, nonlinear and nonstationary characteristics make it hard to be modeled accurately with conventional monitoring methods. In this paper, local dynamic broad kernel stationary subspace analysis (Local-DBKSSA) is developed to improve the monitoring performance. Faced with complex dynamic nonlinear characteristics, a single model is considered to be unable for accurate representation. Thus, dynamic broad nonlinear features established by time shift and multi-kernel projection are adopted from more perspectives. Subsequently, the above features are integrated into stationary subspace analysis (SSA) to estimate stationary projections from time-varying data. In order to reduce the impact of large fluctuations and improve fault detection capability, a local statistic is further proposed. The effects of nonstationary characteristic on monitoring capability and the excellent performance of the local statistic are also theoretically analyzed. Finally, a case study based on actual BFIP data presents that the proposed method can discriminate between normal and sample faults more accurately and timely, and has better robustness to nonstationary perturbations under normal conditions by providing fewer false alarms.

Soft sensor of silicon content in hot metal is challenging due to the nonlinearity, dynamics, and non-Euclidean coupling relationship between process variables. The inter-variable coupling relationship widely exists in the ironmaking process. Nevertheless, they are generally overlooked in modeling. To solve this issue, this paper proposes a novel temporal graph convolutional network with supervised graphs (TGCN-S). Different from traditional soft sensor methods, TGCN-S explicitly models the inherent non-Euclidean and irregular coupling relationships in the ironmaking process. First, a graph structure learning module is designed, which can adaptively learn potential inter-variable relationships from data. This module is embedded in TGCN-S to learn the supervised graph structures in an end-to-end manner. Second, a novel methodological framework is proposed based on the supervised graph structures, which can collaboratively extract the inter-variable coupling relationships and intra-variable dynamics. In this structure, information between process variables is selectively aggregated while taking into account the dynamics within variables. Finally, experiments based on the real ironmaking process demonstrate the effectiveness of the proposed method.

Broad learning system (BLS), a tri-layer feedforward neural network, has gained widespread recognition for its exceptional scalability and computational efficiency. However, BLS and its derivatives encounter several challenges: 1) overlooking the uncertainty introduced by numerous nonlinear random mappings; 2) failing to cope with the misalignment of model inputs with the output sampling rate; 3) lack of attention to nonstationary scenarios; and 4) absence of theoretical optimization for incremental learning. To overcome these obstacles, we propose a regression modeling and anomaly detection scheme rooted in a temporal kernel stationary BLS (TKS-BLS). We first create a nonlinear kernel broad representation (NKBR) extraction strategy, providing a robust nonlinear foundation for random feature mapping via kernel technology. Following this, we probe the mechanism of temporal matching between model inputs and outputs through a temporal alignment parameter, interpretable under a latent variable relationship. In the integration phase, we establish a Kullback-Leibler divergence objective function to facilitate the capture of stationary relationships within time-series data, in conjunction with the regression error. Subsequently, a double-loop parameter optimization algorithm and an independent incremental learning mechanism are put forth, both backed by comprehensive theoretical analyses. Our method’s superiority is thoroughly confirmed by experimental outcomes from extensive case studies across seven real ironmaking process datasets.

To ensure the operation safety of complex industrial processes, faults that occur in the process should be detected and identified in time to avoid further catastrophic events. Therefore, fault diagnosis plays an indispensable role in the process industry. With the expansion of the production scale of industrial processes, the structural attributed graph is suitable for describing the process data structure due to the complicated interactions between sensor measurements. Graph convolutional networks (GCN) can identify and capture the relationships between industrial process data in non-Euclidean space by taking graph data with topological structure as input. In this paper, a novel fault diagnosis method based on a multi-channel dynamic graph convolutional network (MDGCN) is proposed. Different from traditional graph convolutional networks, the proposed MDGCN assigns different weights to the nodes of the graph. Thus, more useful information about process dynamics is extracted. Particularly, the strategy of multiple isomorphic graph channels is developed to learn feature representations of process data from different levels for fault diagnosis. The capability and efficiency of the proposed MDGCN-based fault diagnosis method are demonstrated through an industrial benchmark of the Tennessee Eastman process and a real blast furnace iron-making process.

Blast furnace iron-making process (BFIP) is one of the most critical procedures in the iron and steel industry where timely detection and accurate classification of faults have always been of core focus. However, the coupling effects of system’s nonlinear and nonstationary characteristics often cause process consistent underlying information to be buried, allowing accurate extraction to be a significant challenge. This also complicates the development of BFIP fault diagnosis model. Therefore, we propose a novel data-driven joint fault diagnosis strategy that employs regularized mutual kernel analytic stationary subspace analysis (RMK-ASSA) and deep broad stationary kernel network (DBSKNet) to eliminate this interference. To develop this method, we first construct an RMK-ASSA approach to address the poor modeling accuracy caused by standard analytic stationary subspace analysis (ASSA)’s inability to handle complex process nonlinearity. Global and local kernels are utilized to account for multiple nonlinearities in BFIP data. The weight of different nonlinear data is calculated by regularized principal component analysis, and the main information is imported into ASSA to obtain more robust and accurate modeling results by eliminating the interference of redundant noise. Subsequently, we design a DBSKNet-based classifier to implement the fault diagnosis task. This network further considers the nonlinearity by boosting kernel structure in depth and width while distinguishing the respective contributions of different kernels to fault diagnosis results. Finally, a double-layer loop parameter optimization algorithm is used for optimizing. Simulated cases and practical BFIP tests validate that RMK-ASSA eliminates the negative impact caused by nonstationary data and that the proposed joint fault diagnosis strategy outperforms other methods. Note to Practitioners —BFIP’s nonlinear and nonstationary coupling properties pose unique challenges in eliminating distractions, constructing fault classifiers and accurately detecting process anomalies. To tackle these challenges, this paper proposes a joint fault diagnosis strategy based on RMK-ASSA and DBSKNet. RMK-ASSA effectively estimates nonlinear consistent features, while DBSKNet mines rich deep nonlinear information, accurately distinguishing variations in BFIP data under different working conditions. Experimental results demonstrate that this data-driven strategy can perform high-quality fault diagnosis, enabling field engineers to execute operations efficiently.