Safety monitoring of power operations in substations is crucial for accident prevention. However, traditional methods such as wearable devices and video surveillance suffer from limitations including high costs and reliance on lighting conditions. A solution that integrates millimeter-wave radar with deep learning ensures operational compliance by high precision gesture detection. This paper proposes RoPEHAR, a human activity recognition system based on millimeter-wave radar, which combines traditional transformer architecture with rotary positional encoding, specifically designed for human posture recognition in indoor industrial environments. To reduce interference from the coupling of human and instrument signals and noise in electrical scenarios, RoPEHAR introduces a hybrid filtering pipeline that combines hierarchical SNR denoising with enhanced DBSCAN clustering to accurately segment the point cloud data of limbs and instruments. The core innovation lies in the introduction of a spatiotemporal Informer, Roformer. It enhances the 3D-space vector information about data points through dynamic rotary positional encoding. This system effectively models limb motion trajectories. Experiments demonstrate that RoPEHAR achieves high-precision performance, reaching an accuracy of 95.8%, enabling real-time and reliable activity recognition for substation safety monitoring.

Accurate and scalable indoor localization for unmanned aerial vehicles (UAVs) is essential for Internet of Things (IoT) applications such as autonomous logistics, infrastructure inspection, and emergency response in GPS-denied environments. However, traditional methods often struggle with cost, deployment complexity, and sensitivity to environmental dynamics, limiting their practicality for large-scale IoT scenarios. This paper presents a method in which Channel State Information (CSI) from low-cost IoT sensors enables robust, device-free 3D UAV localization while optimizing accuracy, sensor adaptability, and data efficiency. We propose CiUAV, leveraging CSI captured by ESP32-S3 sensors, with a Robust CSI Signal Enhancement (RCSE) framework integrating Dynamic AGC Compensation (DAC) and Adaptive Noise Suppression and Outlier Removal (ANSOR), alongside a Sensor-in-Sample (SiS) multi-objective optimization model for adaptive multi-sensor fusion. Experimental evaluations in realistic indoor settings achieve a 3D root mean squared error (RMSE) of 0.2659 meters, outperforming baselines by up to 35% in accuracy and 50% in data efficiency. CiUAV offers a lightweight, scalable, and infrastructure-compatible solution for future IoT-enabled UAV systems.

Fault diagnosis for quality-related variables is crucial for maintaining product quality and operational efficiency of industrial processes. However, modern industrial processes exhibit inherent complexity, making it difficult to accurately diagnose quality-related fault variables. In addition, traditional methods such as contribution plots lack the ability to quantitatively evaluate fault variables. To overcome these deficiencies, this paper proposes a canonical correlation residual score (CCRS)-based fault diagnosis method. Different from conventional methods, the CCRS model is constructed by means of canonical correlation between quality-related and quality-unrelated process variables, where the influence of noises in quality-related variables is reduced. Furthermore, the performance of the CCRS-based diagnostic method is theoretically analyzed. Necessary and sufficient conditions are derived to quantitatively characterize the score of canonical correlation residuals. Finally, an online implementation of CCRS is developed to iteratively update the CCRS model for quality-related fault diagnosis in nonstationary industrial processes. The diagnosis performance of the proposed method has been validated on an experimental three-tank system and the Tennessee Eastman process.

This article investigates an integrated approach of fault diagnosis and initial alignment of redundant strapdown inertial navigation systems (SINSs) under large misalignment angles. A redundant configuration of four hemispherical resonator gyroscopes (HRGs) and four accelerometers is designed. The parity vector method combined with generalized likelihood ratio test is developed for reliable detection and identification of HRG bias faults. For initial alignment, an analytic coarse alignment provides an initial attitude estimate, which is followed by a precise alignment phase using an unscented Kalman filter (UKF). The UKF is specifically designed to handle the nonlinear error model associated with large yaw misalignment angles. Experimental results demonstrate that the proposed fault diagnosis method effectively identifies HRG faults. Furthermore, comparative studies show that while the UKF and extended Kalman filter yield similar performance for small misalignment angles, the UKF achieves significantly superior alignment accuracy, especially under large yaw misalignment angle. This integrated approach enhances system reliability and navigation precision, which achieves a 100% detection rate for the tested bias faults and reduces the yaw error from$20^{\circ }$to below$0.35^{\circ }$.

Ensuring personnel safety during power station inspections is a critical yet challenging task due to inherent hazards in such environments. Traditional monitoring methods, including wearable devices and video surveillance, suffer from user discomfort, limited visibility, and high deployment costs. To overcome these limitations, this article proposes PowerHAR, a device-free framework for recognizing power inspection activities based on WiFi channel state information (CSI) acquired from custom-designed ESP32 internet of things (IoT) sensors. PowerHAR introduces a spatio-temporal feature distribution-based power operation recognition network, comprising a transformer-based preprocessing module capable of effectively handling variable-length CSI sequences, and a spatio-temporal extraction module that integrates convolutional operations with multihead self-attention mechanisms for comprehensive feature fusion. By leveraging mutual CSI sensing among distributed sensors, PowerHAR provides robust and accurate recognition of power inspection activities without requiring additional hardware infrastructure. Experimental validation demonstrates that PowerHAR significantly surpasses existing baseline methods, confirming its high reliability and practicality in safety-critical industrial scenarios.

Fault diagnosis is crucial for ensuring the safety and reliable operation of train traction drive systems. Traditional methods primarily depend on expert experience and historical fault records for analysis. However, as knowledge sources and types expand, existing approaches struggle to integrate multi-source heterogeneous knowledge effectively, which compromises the accuracy of fault diagnosis and the formulation of maintenance decisions. To address these challenges, this paper proposes a multisource knowledge retrieval-augmented large language model (LLM)-based fault diagnosis method. First, a multisource retrieval mechanism is introduced, which uses vector databases and knowledge graphs to extract highly relevant unstructured text and structured knowledge. Next, a reranking model refines the retrieved information, which ensures that the large language model accesses high-quality reference knowledge. Finally, prompt templates integrate optimized textual and structured knowledge that guide the LLM to generate accurate and interpretable fault diagnosis responses. Experimental results show that the proposed method significantly improves the accuracy and reliability of fault diagnosis for traction drive systems while improving the applicability of large language models.

The occlusion problem has consistently posed a significant challenge in the field of segmentation. Most existing segmentation methods require additional annotations and fail to capture the contour information of occluded regions, thus not truly addressing the occlusion issue. Although segmentation tasks involving particle objects also suffer from occlusion problems, the homogeneity of particle objects offers new possibilities for overcoming this challenge. In this paper, we propose an occlusion segmentation framework for particle objects that does not require additional annotations. This framework only necessitates instance-level segmentation labels to obtain complete contour information of particle objects, including occluded regions. First, we decompose the occlusion segmentation task into a generic instance segmentation task and an occlusion repair task for occluded objects. Then, in order to train the occlusion repair model with only instance segmentation-level labels, we quantitatively analyze the occlusion phenomenon, including the mathematical descriptions of occlusion relationships, degrees, and distributions. Next, we geometrically transform and layer overlay the unobscured samples to construct occlusion samples containing labeling information of the occluded regions. These sample sets are used to train a generative model that predicts the contour information of occluded regions. Finally, we fine-tune or post-process the pre-segmentation model with the particle objects containing restored complete contour information to achieve the final occlusion segmentation. We conducted extensive ablation experiments on both the ore-particle dataset and publicly available cell-particle datasets. The experimental results validate the effectiveness, accuracy, and generalizability of our method. Note to Practitioners—Particle segmentation has been faced with the occlusion problem. In this paper, inspired by the similarity between particle objects, we propose a self-supervised occlusion segmentation framework that does not require additional annotation of occlusion layers. Our approach requires only instance segmentation level annotation without more complex additional manual annotation, which is crucial for practical applications. In addition, we decouple the complex occlusion relation modeling into a binary classification problem without knowing precisely the occlusion hierarchy between particles, which further reduces the difficulty of practical applications. Then, we also propose shading transformations to characterize the inter-particle shading distribution to construct shading sample sets from existing samples. Finally, we use these learned and constructed occlusion sample sets to pre-train the generative model for regenerating the occluded objects to complete the final occlusion segmentation. Although in a generic segmentation task, our approach may have some limitations because the segmented objects may not have an apparent similarity. However, our approach using self-supervision and the objects’ properties provides valuable ideas for solving the occlusion problem. In the future, we will solve the occlusion problem regarding the properties of each class of objects rather than just considering the similarity among particle objects.

In this article, a correlation-aided neural network (CANN) framework is proposed for distributed process monitoring of large-scale industrial automation systems. First, to model the nonlinear data relationship of multiple subsystems, a feature space is learned for each subsystem by means of neural network-based nonlinear projection. Second, different from conventional distributed monitoring methods, a distributed learning method and a novel objective function are proposed, where the influence of uncertainties is decreased by considering the correlation of variables among multiple subsystems. Third, a process monitoring method is proposed based on the CANN framework. The impact of the fault is analyzed, and the optimal monitoring performance is achieved by considering the correlation of all subsystems. To further reduce communication overhead among multiple subsystems, a tradeoff between monitoring performance and communication overhead is made based on gradients of the CANN model. Theoretical analysis demonstrates the superiority of the proposed method. The case studies on a multizone heating, ventilation, and air-conditioning system and the Tennessee Eastman benchmark process are given to evaluate and compare the effectiveness of the proposed method.

The sintering process is crucial for the preparation of ternary cathode materials (TCMs), which need to be precisely monitored to ensure the production of high-quality products. Nevertheless, the sintering process of ternary cathode materials (TCMs-SP) is marked by a prolonged duration, the high dimensionality of process variables, and spatio-temporal correlation coupling, which poses challenges when using traditional centralized monitoring methods. To this end, a novel distributed monitoring method suitable for TCMs-SP is proposed in this article. First, the sintering process is divided into different subsystems based on the material manufacturing process. Then, during the establishment of the local monitoring model, using the low-dimensional strong correlation information transmitted by other subsystems, the Gaussianity preserved deep neural network-aided canonical correlation analysis (CCA) is utilized to extract the Gaussianized nonlinear dynamic features. On this basis, the monitoring statistics of the input and output observation space are developed via CCA, and a monitoring strategy is specially designed for TCMs-SP. Finally, the industrial application verifies that our proposed approach offers superior monitoring compared to existing methods. It allows for rapid localization of faults and enhances troubleshooting efficiency in the sintering process.

Ensuring the accurate recognition of blast furnace (BF) conditions is crucial for maintaining the stability of blast furnace ironmaking process (BFIP). However, many research frequently overlooks the comprehensive utilization of the closely correlated multisource heterogeneous data (MHD) of BFIP, such as time series and images, resulting in an incomplete understanding of BF conditions and an unsatisfactory accuracy of condition recognition. Therefore, this article proposes a novel BF condition recognition method by incorporating both time series and images of BFIP. First, a data alignment model using the Gaussian functions is proposed to align time series with images. Then, an innovative Time Series—Image jointly driven deep neural Network (TSIN) is established to extract and fuse features from MHDs. Subsequently, a dual-layer residual-connected module is designed to capture the correlations between MHDs. Furthermore, a stability evaluation metric is devised to quantify the stability of TSIN during the model training process. Industrial experiments demonstrate that the proposed method, with an average recognition accuracy of 96.99%, could effectively recognize abnormal conditions, such as channeling, hanging, slipping, and collapsing, providing practical guidance for on-site workers to monitor and regulate the BFIP.

Confidence calibration in classification models is a vital technique for accurately estimating the posterior probabilities of predicted results, which is crucial for assessing the likelihood of correct decisions in real-world applications. Class imbalance data, which biases the model's learning and subsequently skews predicted posterior probabilities, makes confidence calibration more challenging. Especially for underrepresented classes, which are often more important and tend to have higher uncertainty, confidence calibration is more complex and essential. Unlike previous surveys that typically separately investigate confidence calibration or class imbalance, this article comprehensively investigates confidence calibration methods for deep learning-based classification models under class imbalance. First, the problem of confidence calibration under class imbalance data is outlined. Second, this article explores the impact of class imbalance data on confidence calibration in theory, providing some explanations for empirical findings in existing studies. Third, this article reviews 60 state-of-the-art confidence calibration methods under class imbalance data, divides these methods into six groups according to method differences, and systematically compares seven properties to evaluate their superiority. Then, some commonly used and emerging evaluation methodology are summarized, including public datasets and evaluation metrics. Subsequently, this article performs necessary comparative experiments to provide better guidelines and insights to the readership. Finally, we discuss several application fields and promising research directions that serve as a guideline for future studies.

The nonlinear and dynamic nature of complex industrial processes presents a significant challenge for monitoring incipient faults. To this end, this article proposes a novel deep dynamic latent variable model called dynamic inner canonical variate network (DiCVNet). The developed DiCVNet, which is in an end-to-end learning framework, consists of a dual convolutional autoencoder (DuCAE) and an autoregressive (AR) module. First, the DuCAE architecture with an AR module is designed to extract two correlated and self-orthogonal nonlinear dynamic canonical variables (CVs) from past and future datasets for tiny variation modeling. The AR module is embedded to extract the CVs with consistent dynamics for enhanced dynamic modeling of DuCAE. Then, a new incipient fault monitoring scheme for nonlinear dynamic processes is established. Finally, the performance of the proposed method is verified by two industrial cases, that are, a continuous stirred tank reactor and a multiphase flow process.

A traction drive system (TDS) in high-speed trains is composed of various modules including rectifier, intermediate dc link, inverter, and others; the sensor fault of one module will lead to abnormal measurement of sensor in other modules. At the same time, the fault diagnosis methods based on single-operating condition are unsuitable to the TDS under multi-operating conditions, because a fault appears various in different conditions. To this end, a real-time causality representation learning based on just-in-time learning (JITL) and modular Bayesian network (MBN) is proposed to diagnose its sensor faults. In specific, the proposed method tracks the change of operating conditions and learns potential features in real time by JITL. Then, the MBN learns causality representation between faults and features to diagnose sensor faults. Due to the reduction of the nodes number, the MBN alleviates the problem of slow real-time modeling speed. To verity the effectiveness of the proposed method, experiments are carried out. The results show that the proposed method has the best performance than several traditional methods in the term of fault diagnosis accuracy.

Coverage control describes the optimal deployment problem of multi-agent system with communication sensors, aiming to drive the multi-agent system reach the optimal deployment location. To accomplish coverage tasks, agents are required to communicate with each other. Thus, the connectivity of multi-agent system is a fundamental requirement in case of agent disconnection and disappearance. In this paper, we propose a distributed coverage control algorithm with connectivity optimization. According to the designed cost function, the controller is divided into two parts: coverage controller and connectivity optimization controller, and the method of gradient descent is used to minimize the cost function to get the controller. Finally, the simulation serves to show the effectiveness of the algorithm.

Due to the non-intuitive high-level semantics of fault samples, such as the fault type and operating condition, traditional unsupervised methods are not suitable for fault data augmentation in traction motors. To this end, a new method termed disentangled semantic embedding in variational autoencoder(DSE-VAE) is proposed to learn the interpretable representation of fault samples. In DSE-VAE, a regularization term is constructed by introducing attribute labels to bridge the latent space and semantics. In addition, the mutual information between different attributes is minimized to achieve attribute disentanglement. Extensive experiments are conducted on the hardware-in-the-loop (HIL)real-time platform. Experimental results suggest that the proposed DSE-VAE could learn an interpretable fault sample generation.

Unlike general image segmentation, highly complex particle images have significant challenges in labeling and segmentation due to the information occlusion and texture disturbance. Aiming at the highly under-segmentation problem caused by complex particle image segmentation, this paper proposes a Semi-supervised Hybrid-training Particle Segmentation framework (SHPS) based on skeleton-guided pseudo-labels. First, a pre-trained model is obtained by training a popular segmentation algorithm on partially labeled data. Then, a Background Skeleton-guided Pseudo-label generation algorithm (BSP) is proposed to generate pseudo-labels closer to the ground truth in terms of structural integrity based on coarse segmentation. The final segmentation model is obtained by training a mixed dataset consisting of labeled data and pseudo-labels from another partition on the pre-trained model. The skeleton differences of pseudo-labels and coarse segmentation are added to the loss function. Experimental results show that our method achieves 84.4% accuracy on mIoU with uniform label data distribution, which is 2.1% higher than the accuracy of UNet and reduces the degree of under-segmentation.

Intelligent fault diagnosis of the complex systems has made great progress based on the availability of massive labeled data. However, due to the diversity of working conditions and the lack of sufficient fault samples in practice, the generalization of the existing fault diagnosis methods are weak. To handle this issue, a multichannel domain adaptation graph convolutional network method is proposed. In the proposed network, a feature mapping layer based on convolutional neural network is used first to extract features from input data, which then are transmitted to the graph generator to construct two association graphs. After that, three distributed graph convolutional networks are used to extract the specific and common embeddings from two association graphs and their combination. Meanwhile, to fuse these embeddings adaptively, an attention mechanism is used to learn importance weights. Besides, a domain discriminator is leveraged to reduce the distribution discrepancy of different data domains. Finally, a label classifier is used to output fault diagnosis results. Two experimental studies with different signal types show that the proposed method not only presents better diagnosis performance than existing methods with few samples, but also can extract domain-invariant features for cross-domain under varying working conditions.

Due to the interconnected characteristics between subsystems and the strong correlation within subsystems, the monitoring of plant-wide processes has become a challenging problem, especially for tandem plant-wide processes that exist in various industrial fields, such as petrochemicals, metallurgy, and sewage treatment. In this article, a novel spatio-temporal monitoring method is proposed for the hot strip mill (HSM) process, a typical tandem industrial process. First, the plant-wide process is divided into different subblocks based on the tandem structure. Then, a distributed conditional variational recurrent autoencoder-based process monitoring method is proposed to build the local latent variable model of each subsystem using relevant dynamic features extracted from the previous subsystem. The latent distributions and reconstructed errors are used to design local monitoring statistics for local process monitoring. A global monitoring statistic is established by deep support vector data description to monitor the whole process. Finally, the effectiveness and superiority of the proposed method are demonstrated by a HSM process case, which shows better monitoring performance compared to the existing methods.

In the conventional graph convolutional network (GCN)-based fault diagnosis method, multilayer GCN model is often used for feature extraction. However, the application of multilayer GCN will encounter oversmoothing problem, and thus reduce the diagnostic performance. Therefore, the oversmoothing relief GCN (OsR-GCN) method is proposed. Specifically, two association graph construction methods, namely the Euclidean distance (ED)-based method and the structure analysis (SA)-based method, are first introduced. Then, the constructed graph and measurements are input to the OsR-GCN model, in which a weight coefficient is proposed to relieve the oversmoothing problem. Next, an improved particle swarm optimization algorithm is introduced to find the optimal weight coefficient. Finally, the proposed method is applied to diagnose the pulse rectifier faults in a hardware-in-the-loop simulated traction control system of high-speed trains. The achieved results show that the proposed method outperforms the existing fault diagnosis methods.

Recently, canonical correlation analysis (CCA) has been explored to address the fault detection (FD) problem for industrial systems. However, most of the CCA-based FD methods assume both Gaussianity of measurement signals and linear relationships among variables. These assumptions may be improper in some practical scenarios so that direct applications of these CCA-based FD strategies are arguably not optimal. With the aid of neural networks, this work proposes a new nonlinear counterpart called a single-side CCA (SsCCA) to enhance FD performance. The contributions of this work are four-fold: 1) an objective function for the nonlinear CCA is first reformulated, based on which a generalized solution is presented; 2) for the practical implementation, a particular solution of SsCCA is developed; 3) an SsCCA-based FD algorithm is designed for nonlinear systems, whose optimal FD ability is illustrated via theoretical analysis; and 4) based on the difference in FD results between two test statistics, fault diagnosis can be directly achieved. The studies on a nonlinear three-tank system are carried out to verify the effectiveness of the proposed SsCCA method.

Deep-neural network-based fault diagnosis methods have been widely used according to the state of the art. However, a few of them consider the prior knowledge of the system of interest, which is beneficial for fault diagnosis. To this end, a new fault diagnosis method based on the graph convolutional network (GCN) using a hybrid of the available measurement and the prior knowledge is proposed. Specifically, this method first uses the structural analysis (SA) method to prediagnose the fault and then converts the prediagnosis results into the association graph. Then, the graph and measurements are sent into the GCN model, in which a weight coefficient is introduced to adjust the influence of measurements and the prior knowledge. In this method, the graph structure of GCN is used as a joint point to connect SA based on the model and GCN based on data. In order to verify the effectiveness of the proposed method, an experiment is carried out. The results show that the proposed method, which combines the advantages of both SA and GCN, has better diagnosis results than the existing methods based on common evaluation indicators.

Concrete cracks are one of the most apparent indicators for possible structural deterioration and need to be periodically inspected. However, for current image-based automated crack inspection techniques, accurate and detailed crack quantification and assessment remain a challenging task. Most of these techniques require high-quality input images, which may be difficult to ensure in practice. Besides, simply merging crack detections from multiple images to generate a large crack map may result in an inaccurate outcome for crack severity assessment. In this article, a novel crack quantification framework is proposed to identify complete crack geometric properties utilizing a set of unordered inspection images. To realize this, cracks in images are detected by an instance segmentation convolutional neural network. Subsequently, the crack segmentations from multiple separate images are systematically aggregated through voxel-based reconstruction and Bayesian data fusion. This framework outputs a crack model that can retrieve accurate geometric properties of each crack segment by recognizing the crack's inherent branching patterns. The capability and performance of the proposed crack quantification framework are validated on cracked concrete specimens in a laboratory setting. Also, a field test on a cracked concrete wall was carried out using images captured by a UAV to demonstrate the efficacy of the proposed framework in practical conditions.

The pulse rectifier of a traction system in locomotives and electric multiple unit (EMUs) is usually vulnerable to performance degradation and faults due to uncertain factors, such as vibrations, aging, electromagnetic interferences. In order to ensure adequate redundancy and isolation measures in time to avoid fault propagation in traction systems, a real-time fault diagnosis method for sensors and IGBTs of the impulse rectifier is proposed, which lays a foundation for the redundant design of traction systems. Once a fault is detected in this paper, which can immediately act to the detected faults and take effective remedial measures to avoid the failure of the whole system. It is based on the structural analysis of the traction system whose structural model of interest will be established. Meanwhile, the structural model is evaluated and optimized according to the analytical relation model under various fault conditions, and the minimum structural overdetermined sets (MSOs) are obtained based on the optimized model, which can be used to isolate all faults. Using the MSOs, the redundancy relationship is deduced and the sequence residuals are generated. Afterward, the cumulative sum (CUSUM) algorithm is used for diagnosis decision making. The effectiveness of the proposed method is finally verified on a hardware-in-loop test platform, which can accurately simulate a traction system. It shows that the proposed method can achieve both good feasibility and high accuracy.

Multivariate analysis is an important kind of method in process monitoring and fault detection, in which the canonical correlation analysis (CCA) makes use of the correlation change between two groups of variables to distinguish the system status and has been greatly studied and applied. For the monitoring of nonlinear dynamic systems, the deep neural network-aided CCA (DNN-CCA) has received much attention recently, but it lacks a general definition and comparative study of different network structures. Therefore, this article first introduces four deep neural network (DNN) models that are suitable to combine with CCA, and the general form of DNN-CCA is given in detail. Then, the experimental comparison of these methods is conducted through three cases, so as to analyze the characteristics and distinctions of CCA aided by each DNN model. Finally, some suggestions on method selection are summarized, and the existed open issues in the current DNN-CCA form and future directions are discussed.

Threats of cyberattacks have penetrated from disclosing critical user information to destroying/manipulating industrial control systems. Study on data security during network transmission has raised increasing attention in the systems and control community, which is found very necessary and timely in the context of Industry 4.0. In most existing approaches, the protection of the transmitted data from eavesdropping attacks and the detection of malicious integrity attacks are usually carried out separately. In this study, an integrated data-driven framework applicable at the control level is proposed to deal with secure transmission and attack detection simultaneously. In the framework, a secure correlation-based encryption/decryption approach and a trustworthiness judgement approach are proposed. Comprehensive discussions are made regarding the analysis of the sensitivity to attacks, the introduced time delay, and the design degree-of-free. Executable algorithms are presented, corresponding to which hardware is modularized and can work standalone independent from the configuration of the monitoring and control systems or any third-party authentication agencies. Evaluation results on a simulated two-area frequency-load control power grid system are provided to show the effectiveness and performance of the proposed approaches.

This article proposes a new data-driven health monitoring method, which uses multiobjective optimization and stacked autoencoder based health indicator. Specifically, the proposed method proposes an improved nondominated sorting genetic algorithm-II (NSGA-II) to perform multiobjective optimization on a large number of candidate features extracted from the sensor measurements. Then, a stacked autoencoder model is used to construct health indicators from the selected features. In the improved NSGA-II algorithm, the optimization goals of feature selection are defined as the minimum gap of health indicators between different states and the number of features. Comparisons between the proposed method and the state-of-the-art methods on simulation experiments show that the proposed method can accurately identify the status of the equipment and effectively limit the complexity of the diagnostic model.

In modern large buildings, the chilled water system regulates the indoor temperature through a series of heat exchanges. This article studies the optimal control problem of the chilled water system and proposes an ensemble learning method for the cooling load prediction under different operation conditions with imbalance sample distribution. A new control strategy is afterwards developed for optimal selection of the process control inputs that guarantee the demand for cooling load with a lower energy consumption. The optimal control strategy is also learned in real time using a cloud edge terminal form, which can be used for big data modeling and increase the effectiveness of the system response. The proposed method is applied to a real high-rise building, and the results show a significant improvement in the proposed prediction model and the optimal control strategy, compared to the state-of-the-art methods. Regarding manual operation, the control strategy decreased the energy consumption by 5.59%, and, on average, 35 645 kWh of electric energy per month was saved.

Demagnetization, as one of the most frequent faults, has great influence on the performance of (permanent magnet synchronous motor) PMSM. However, the motor usually runs in nonstationary conditions, that brings great challenge to the effective diagnosis of demagnetization fault. This paper presents a new methodology of Shift-invariant Dictionary of Dynamic Time Warping (SIDDTW) to diagnose the demagnetization fault under nonstationary conditions. Firstly, according to the characteristics of current signal under demagnetization fault, the shift-invariant dictionary is constructed. Then, Matching Pursuit (MP) is used to represent the current signals that collected from the running process of PMSM, and then the sparse coefficient series are obtained. Finally, the Dynamic Time Warping (DTW) method is used to calculate the sparse coefficient series distance between the test data and the database which build in the training process. In this step, the nearest distance is matched, and corresponding operation state is recognized as the final diagnosis result. The results show that the presented method has good adaptability when dealing with nonstationary conditions both on the Simulink platform and the real-time simulation platform.

In the manufacturing process of ternary cathode material, batching, mixing, loading and sintering procedures play decisive roles on product quality, and data generated by these procedures are collected by the manufacturing process of ternary cathode material. However, data of the process are heterogeneous and have different properties, and sampling of key quality variables is difficult. On this basis, on-line evaluation on product quality is hard. In this paper, a data fusion based online product quality evaluation method for the ternary cathode material is proposed. Data is collected from an enterprise manufactures ternary cathode material. Then, to establish the model between the extracted data features and the key quality variables such as particle size and surface-free lithium content, a semi-supervised double-weighted probabilistic principal component regression is proposed. After that, a distance cost index is brought to cluster and grade the two quality variables, which are predicted by the semi-supervised model. Therefore, the on-line evaluation of product performance is achieved by setting rule table in accordance with production experience. Finally, based on data collected from an enterprise manufactures ternary cathode material, the proposed method is verified to be effective and accurate.

In this paper, a new data-driven fault detection method based on distributed canonical correlation analysis (D-CCA) is proposed to address the plant-wide process monitoring problem. This paper focuses on the distributed plant-wide processes. The core of the proposed method is to reduce uncertainties using correlation information from the neighboring nodes. Furthermore, the cost of the data transmission between network nodes is also reduced by the D-CCA algorithm. When the proposed method and the existing methods are compared using the Tennessee Eastman benchmark process, the false alarm rate, fault detection rate, and the detection delay are comparable. This suggests that the proposed method is feasible.

Hot spots are common abnormalities in photovoltaic (PV) energy systems. Their presence can potentially cause damage to PV modules, such as performance degradation or even unexpected fire to PV energy systems. By sufficiently mining the information hidden in the test data collected from PV modules, this paper develops a space-to-space projection method, which at its core is a linear approach via preserving the locally geometrical structure with respect to time series. Based on the nonlinear model of PV modules established via the proposed projection, data-driven detection of hot spots in PV energy systems can be directly achieved with three key advantages: 1) its implementation does not depend on any mathematical model or physical knowledge of PV energy systems; 2) it is of high-computational efficiency especially in the online detection phase; and 3) it can capture the dynamic characteristic because the local structure of samplings regarding time is given sufficient consideration. The effectiveness and feasibility of the proposed approach are first presented by theoretical analysis and, then, convictively demonstrated via 15 sets of hot spot experiments on practical PV modules.

This paper presents an adaptive data-driven fault detection method for dynamic processes. In this method, the vector ARX model is used to model the dynamic process in a data-driven fashion. Then, the adaptive method is developed by means of the incremental and decremental algorithms. The performance and effectiveness of the proposed approach are demonstrated with a numerical case study and an experimental continuous stirred tank heater. The detection results show that the effectiveness of the proposed method.

Fisher discriminant analysis (FDA) is a widely used tool for fault diagnosis. In addition, many modifications have also been proposed recently in the literature in order to overcome certain limitations of the traditional FDA method. However, the incipient fault diagnosis problem is not well handled by traditional FDA and its variants. In this paper, through the introduction of smoothing techniques, a new method called smoothed FDA (SFDA) is proposed to enhance the fault diagnosis performance for incipient faults. Fault diagnosability analyses of the FDA and SFDA approaches are carried out and compared with each other. It is pointed out through theoretical analysis that SFDA is superior to FDA in terms of incipient fault diagnosis. Simulation studies on a continuous stirred tank reactor process are used to demonstrate the effectiveness of the SFDA method, in comparison with traditional FDA.

The problem of fault diagnosis for static processes has been well studied over the last decades. However, fault diagnosis methods have rarely considered processes subject to unknown periodic disturbances. Using the well-established orthogonal function technique, this paper proposes a data-driven fault diagnosis method to deal with this challenge. The basic idea is to first design the orthogonal functions, and then identify the unknown weighting parameters. By removing the influence of the periodic disturbances, the residual signal can be obtained. Then, the fault detection problem is solved by monitoring the change of the residual signal. The performance and effectiveness of the proposed approach are demonstrated with a numerical case study and an experimental study based on a pilot scale, continuous stirred tank heater.