This paper is concerned with the distributed coordination problem for the nonlinear multi-agent system (MAS) over a general digraph, where each agent is a multi-input multi-output system. The existing solutions are limited to the system without inputs coupling and with known, global Lipschitz, or linearly growing nonlinearities. To remove these requirements, we propose an integrated sub-optimal and control strategy for the more general nonlinear MAS. It consists of a fully distributed adaptive gradient optimization algorithm and a set of model-free prescribed performance controllers. Our approach ensures that the outputs of the MAS converge to the arbitrarily small neighborhoods of the optimal outputs; in particular, the reference-tracking performance is allowed to be freely predefined. Besides, the proposed control strategy is notably simple compared to the existing approaches, which typically employ function approximation, parameter identification, or derivative calculation. Finally, the simulation results illustrate the effectiveness and superiority of the proposed approach.

This article investigates the event-triggered anti-disturbance control for multiagent systems (MASs) subjected to multiple disturbances and time-varying transmission delays (TDs). Unlike existing studies that only consider the abrupt changes in parameters or communication topologies, this work employs the dual-Markov jumping processes with a switching signal to describe stochastic behaviors based on a novel mapping technique. The dynamic event-triggered protocol (DETP) is established to reduce communication burdens by incorporating a packet loss schedule (PLS). Additionally, the composite anti-disturbance controllers are developed based on disturbance observers (DOs) and extended dissipative performance analysis. By employing Lyapunov-Krasovskii functional (LKF) and the Finsler lemma, the stabilization conditions of the switched dual-Markov jumping MAS (SDMJMAS) are derived. Finally, the effectiveness of the proposed methods is validated through comparative experiments.

Industrial data collected from similar processes under varying production specifications or monitoring configurations often exhibit structural heterogeneity, particularly in the form of varying feature dimensions. This presents a fundamental challenge for conventional fault detection models, which typically require fixed-length inputs and assume consistent feature spaces. As a result, practitioners are often forced to either discard valuable heterogeneous data or develop separate models for each product configuration, both of which compromise scalability and efficiency. To address this, we propose a novel heterogeneous AutoEncoder (HAE) framework that enables unified representation learning and fault detection on heterogeneous tabular data with arbitrary feature lengths. HAE adopts a Transformer-based encoder–decoder architecture guided by a binary padding indicator mask, which explicitly distinguishes between observed and padded attributes. To further improve discriminability and reduce intersource variation, a triplet-center loss is introduced to align latent representations and enhance interclass separability. We evaluate our method on two real-world datasets: a heavy-plate production process from the steel industry and a cross-project software defect prediction task. Experimental results show that HAE consistently outperforms a wide range of traditional, statistical, and generative imputation baselines or feature truncation method. Ablation studies further verify the effectiveness of each proposed component. HAE provides a scalable and generalizable solution for heterogeneous fault detection, without requiring strong assumptions on data missingness patterns or feature correspondences.

Stochastic configuration networks (SCNs) are incremental randomized feedforward neural networks with universal approximation properties. However, their single-hidden-layer architecture limits the ability to capture slow-varying features and nonlinear dynamics from complex industrial process data. To address this limitation, we propose a novel framework termed deep enhanced stochastic configuration networks (DESCNs). DESCNs incorporate feature layer, whitening layer, and enhancement layer in addition to the conventional input and output layers. Residual-error-based twin structure of feature and whitening layers is designed to adaptively extract nonlinear slow features in a supervised manner, thereby yielding output-relevant slow features representations. The enhancement layer compensates for nonlinear information loss during feature extraction. This design enables DESCNs to effectively capture both slow-varying and nonlinear dynamics of complex process data, integrating temporal relationships with supervised relevance. The convergence of DESCNs is demonstrated, which guarantees the universal approximation property. Experiments on numerical simulations and real-world industrial datasets validate the superior performance of DESCNs.

The Vehicle Routing Problem with Time Windows under Uncertain Travel Times (VRPTW-UT) is a challenging and practically significant combinatorial optimization problem. Although evolutionary algorithms (EAs) have shown potential in solving VRPTW-UT, they often struggle to balance robustness and convergence. Conventional EA approaches evaluate solutions across multiple disturbance scenes and discard those that become infeasible under any scenario. This often leads to the premature elimination of solutions that are only infeasible in a limited number of scenes, hindering the ability to effectively explore the trade-off between robustness and convergence. To address this issue, this paper proposes a Game-Theoretic Evolutionary Algorithm (GTEA) that models the search process as a game between two adversarial components: a perturbation generation part that constructs high-impact uncertainty scenes, and a robustness enhancement part that improves solutions under those critical conditions. This antagonistic process forces the population to evolve toward solutions that possess both high robustness and convergence, so that GTEA can efficiently produce solutions with high robustness and convergence. Extensive experiments on four benchmark datasets demonstrate that GTEA outperforms five state-of-the-art algorithms designed for VRPTW-UT, achieving superior convergence and robustness.

Imbalanced class distribution disrupts the training of a classifier, resulting in biases favoring majority classes. Data oversampling is a common strategy to tackle this issue. However, traditional methods may generate incorrect and unnecessary instances when facing complex data challenges, such as class overlap, small disjuncts, and noise samples. Therefore, there is a need for an oversampling method that can accurately characterize the data distribution. This paper introduces a novel deep generative oversampling approach for balancing the imbalanced tabular data by leveraging diffusion models and Generative Adversarial Networks (GANs). The model comprises a generator constructed from diffusion models and a discriminator with a Noise-Sensitive Auxiliary Classifier (NSAC) and is trained through an adversarial process. The synergy of these two models enhances stability and sample quality compared to GANs, with faster sampling speed and better conditional generating ability than diffusion models. In experimental validation across 22 real-world datasets, our method consistently outperforms six counterparts regarding Accuracy, F1-score, and MCC for binary and multi-class scenarios. Notably, our approach enhances classifier accuracy for minority classes while maintaining a high level for the majority class, a facet often compromised by other algorithms.

In aerial load transportation scenarios, such as logistics and emergency rescue operations, the load mass and cable length are often unknown in advance, posing significant challenges for the control of quadrotor uncrewed aerial vehicle (UAV) slung-load systems. This article addresses trajectory tracking and anti-swing control of quadrotors under unknown load mass and cable length. First, the system’s dynamic model is established using 3-D force decomposition and the Newton–Euler method. Then, based on the system’s cascade structure, a novel adaptive fixed-time three-loop control scheme is proposed, including the quadrotor’s position and attitude controllers as well as the load’s swing angle controller. Unlike existing asymptotic convergence methods, the approach guarantees fixed-time convergence regardless of initial states, ensuring rapid trajectory tracking and effective suppression of load swing. Notably, the controller integrates adaptive parameter estimation laws that enable real-time online estimation without prior knowledge of system parameters. Moreover, adaptive compensation terms are embedded to enhance robustness against external disturbances. Theoretical analysis guarantees the fixed-time stability of the closed-loop system, while comparative simulations validate the effectiveness and advantages of the proposed method.

Brain-Inspired computing has attracted growing interest in neuromorphic engineering for replicating brain mechanisms to improve artificial neural networks. Although multi-task learning (MTL) outperforms single-task approaches, existing brain-inspired models struggle with knowledge transfer between tasks. This paper introduces B-MRML, a bio-inspired modular reservoir architecture for liquid state machines (LSMs), which enables multi-task learning through specialized blocks inter-connected through recurrent pathways to facilitate cross-task knowledge sharing. To optimize the reservoir structure efficiently, an evolutionary neural architecture search (ENAS) strategy is employed with surrogate models to reduce computational costs. Experimental results confirm that B-MRML significantly enhances MTL performance in brain-inspired systems, offering a novel paradigm for advancing neuromorphic computing research.

The operating conditions in complex industrial processes are often dynamic, unpredictable, and difficult to label, leading to significant challenges in predicting industrial quality indices. To address these challenges and improve prediction accuracy under Multiple Operating Conditions (MOC), we propose a Coarse-Fine Meta-learning Framework (CFMF). Initially, multiple Coarse Models are established using historical operating condition data. We then introduce a MOC-Dynamic Time Warping (DTW) strategy, which utilizes small-batch data from new operating conditions to identify similar time-series characteristics from historical conditions. These similar conditions data are used to train a meta-learning model for the Coarse Models based on Stacking, ultimately resulting in a fine model for quality index prediction. In industrial experiments, we compare the CFMF with classical multi-model learning strategies, and the results demonstrate that the proposed CFMF achieves superior prediction performance on the target domain test set.

Search and rescue (SAR) assumes primary focus during the post-disaster response phase. In recent years, the rapid advances in autonomous aerial vehicle (AAV) target detection technology have opened up new possibilities for SAR operations. However, the images captured by AAV exhibit considerable variation as they dynamically operate at different altitudes, posing challenges for rescue teams in identifying inconspicuous rescue targets. Furthermore, rescuers can hardly trust a detection model with an opaque decision-making process. To this end, we propose an explainable Region of Interest (RoI) attention-based AAV target detection network (RAXNet) capable of detecting small rescue targets with visual explanations. In this model, we first adopt path aggregation network (PANet) as the neck to extract features from different scales. Then, a RoIAttention module is designed to enhance the small target features while providing visual explanations. Specifically, we employ the RoIAlign and nonmaximum suppression methods to obtain region proposals of small targets in the low-level feature layer, followed by an attention-based feature enhancer to focus the extracted region proposals. By combining the enhanced features with the original ones, RAXNet can detect inconspicuous rescue targets and offer corresponding visual explanations, improving model credibility and facilitating rescue efficiency. Finally, we build an AAV visualization system to help rescuers assess disaster sites in real time. The effectiveness and explainability of the proposed method is demonstrated on the VisDrone DET benchmark and RescueNet datasets.

Air transport technology is vital for emergency and transportation tasks, such as disaster relief and crop transportation. This paper proposes a predefined time controller (PTC) for the quadrotor uncrewed aerial vehicle (UAV) transportation system (QUTS) to achieve rapid trajectory tracking while suppressing payload swing under strong coupling and external disturbances. A dynamic model of the QUTS is developed using the Newton-Euler method, and a nonlinear cascade control scheme is established with three loops: an outer loop for position control, a middle loop for payload swing angle control, and an inner loop for attitude control. The boundedness and predefined-time convergence of the QUTS are proven using the Lyapunov method. Compared to the asymptotically stable controller, the fixed-time adaptive controller, and the no-swing angle controller, on one hand, unlike traditional controllers that exhibit longer and condition-dependent settling times, the PTC ensures predefined-time convergence, leading to rapid system stabilization. On the other hand, the PTC achieves higher trajectory tracking accuracy and superior anti-swing performance. Note to Practitioners—In critical missions such as emergency rescue, rapid response and precise control of aerial transportation system are crucial. Unexpected payload swing can lead to safety incidents, and external disturbances further challenge stability. Considering the underactuated nature of the QUTS, the swinging motion of the payload cannot be directly controlled. To address these issues, this paper presents a cascade control scheme featuring predefined time convergence, comprising an outer loop for quadrotor position control, a middle loop for payload swing angle control, and an inner loop for quadrotor attitude control, effectively enhancing the system’s stability and disturbance rejection capabilities. Rigorous mathematical analysis confirms the closed-loop stability of the system, and numerical simulations validate its effectiveness in achieving fast convergence, accurate trajectory tracking, and enhanced anti-swing performance. Future research will focus on applying this control scheme to real-world aerial transportation tasks, particularly in the field of emergency rescue.

This article is concerned with the memory-based event-triggered leader-following dissipative fault-tolerant consensus (LFDFTC) problem for the nonlinear multi-agent systems (NMASs) with semi-Markov switching topologies subject to the generally uncertain semi-Markov (GUSM) jumping process. Unlike the existing event-triggered (ET) consensus results, the dynamic memory event-triggered mechanism (DMETM) and memory-based distributed fault-tolerant (FT) controllers are designed to reduce the ET times. By constructing a general mode-dependent Lyapunov-Krasovskii functional (LKF) and strictly$(\bf {\mathcal {R,Q,T}})-\boldsymbol {\gamma }$dissipative analysis, the dissipative FT consensus conditions of NMASs are derived in this paper. Finally, three actual physical systems are utilized to verify the validity of the proposed method. Note to Practitioners—Owing to the complexity of engineering environment, the consensus control issue of NMASs has attracted widespread attention. Nowadays, the consensus control of NMASs is generally utilized in diverse fields, such as multi-vehicle coordination, smart grids, and unmanned aerial vehicle formation. However, for the electronic device in practical applications, the channel bandwidth is limited due to power and energy constraints, and it is difficult for the fixed communication topologies and traditional periodic sampled-data control method to cope with these unexpected situations. Therefore, the LFDFTC issue for the NMASs with GUSM switching topologies is investigated by adopting DMETM and memory-based distributed FT controllers in this paper. In addition, the proposed FT consensus control methods with prescribed dissipative performance are applied to multiple vehicles time-invariant formation, Chua’s circuits synchronization, and multiple manipulators consensus.

Early faults, with their small amplitudes and slow variations, are easily masked by noise or process trends, making detection difficult. Also, observer-based methods struggle with timely and accurate early fault detection. To address these challenges, a novel observer-driven temporal graph convolutional network (OD-TGCN) is proposed for early fault detection of control systems subject to semi-stationary periodic input disturbances and measurement noise. We employ graph representation to describe the mechanistic relationships and capture temporal features between variables in the control system. First, graph nodes are constructed based on the control inputs, output signals of the control systems and the fault detection residual of the disturbance decoupling generalized proportional-integral observer (DD-GPIO). Then, a method for constructing the adjacency matrix of the graph based on the system matrix is provided. Finally, the output of the temporal graph convolutional network (TGCN) is passed to a Multi-Layer Perceptron (MLP) for graph-based fault detection. The proposed method is applied to a two-wheeled self-balancing robot. Comparative results show that OD-TGCN significantly outperforms DD-GPIO and typical TGCN in early fault detection accuracy. Additionally, OD-TGCN exhibits notable robustness across datasets with different disturbances/noise.

This article investigates the event-triggered multiasynchronous dissipative sliding mode control problem for the gird-connected systems, where the coupled Inductance-Capacitance (LC) oscillators in electrical networks are subject to actuator attacks and external disturbances. To reduce the communication burden, the dynamic event-triggered mechanisms (DETMs) are introduced along with the switching mechanism for multiple topologies. Specifically, the topology switching process is further viewed as a general uncertain semi-Markov (GUSM) jumping process. This jumping process along with the DETM is thus represented by hidden Markov model (HMM). Then the distributed integral-type sliding mode controller is constructed on the top of the HMM. Sufficient conditions for the desired performance of the closed-loop synchronization error system are derived by constructing the mode-dependent Lyapunov-Krasovskii functional (LKF) with extended dissipativity analysis. The numerical simulation of LC oscillators in the single-phase photovoltagic grid interconnection process is conducted to validate the proposed method.

This paper is concerned with distributed time-varying optimization problems for heterogeneous high-order linear multiagent systems (MASs). Compared to the time-invariant case, the outputs of the agents converge to optimal trajectories rather than fixed points. Moreover, the most existing works on time-varying optimization are limited to the first- or second-order MASs. The time-varying and higher-order characteristics of the problem considered in this paper make it particularly challenging. To this end, an integrated optimization and control strategy is proposed, which consists of a novel robust distributed optimizer based on the signum function and a set of tracking controllers. The robustness of the distributed optimizer to reference-tracking errors is guaranteed by the input-to-state stability. The proposed strategy ensures that the outputs of the agents converge to the optimal trajectories with zero optimum-tracking errors. On this basis, the signum function is approximated and replaced by a continuous nonlinear function, resulting in a smooth strategy that effectively eliminates chattering at the cost of bounded optimum-tracking errors. Finally, the proposed approach is applied to the optimal balancing problem for battery energy storage systems to validate its effectiveness.

Most of the current research on dynamic multiobjective optimization problems (DMOPs) assumes that environmental changes can be detectable. However, undetectable changes are frequently encountered in real-world applications, which pose a serious challenge for the existing methods. Because undetectable changes can lead to the failure of change detection techniques, thereby making it difficult to adapt to environmental changes for most algorithms. Therefore, to effectively deal with DMOPs with undetectable changes, this work proposes a dual mutation-based dynamic multiobjective evolutionary algorithm (DM-DMOEA). The proposed DM-DMOEA incorporates the following two main components. First, based on the exploration level of the population, an adaptive selection strategy is proposed, which enables the adaptive identification of individuals for mutation. Second, a dual mutation scheme is developed, utilizing both the polynomial mutation and the Gaussian mutation. These mutation operations are applied on the selected individuals to generate the mutated individuals, allowing for diverse exploration in the search space. After conducting the above two strategies, the population will evolve by the evolutionary criterion of multiobjective optimization. As a result, the algorithm can effectively adapt to undetectable changes in the environment. Comprehensive empirical studies are conducted on different benchmark functions and a real-world application to evaluate the performance of DM-DMOEA. Experimental results have demonstrated that DM-DMOEA is competitive in tracking the Pareto front over time when facing undetectable changes.

Binary optimization assumes a pervasive significance in the context of practical applications, such as knapsack problems, maximum cut problems, and critical node detection problems. Existing techniques including mathematical programming, heuristics, evolutionary computation, and neural networks have been employed to tackle binary optimization problems (BOPs), however, they grapple with the challenge of optimizing a large number of binary variables. In this paper, we propose a dimensionality reduction method to assist evolutionary algorithms in solving large-scale BOPs, which is achieved based on neural networks. The proposed method converts the optimization of a large number of binary variables into the optimization of a small number of network weights, resulting in a significant reduction in search space dimensionality. Crucially, the proposed method obviates the necessity for a training process, which eliminates the requirement for a priori knowledge and enhances the search efficiency. On six types of single-and multi-objective BOPs with up to 10 000 000 variables, the proposed method demonstrates superiority over top-tier evolutionary algorithms and neural network-based methods.

Surrogate-assisted evolutionary algorithms (SAEAs) rely on the infill criterion to select candidate solutions for expensive evaluations. However, in the context of expensive constrained multi-objective optimization problems (ECMOPs) with complex feasible regions, guiding the optimization algorithm towards the constrained Pareto optimal front and achieving a balance between feasibility, convergence, diversity, exploration, and exploitation using a single infill criterion pose significant challenges. We propose an ensemble infill criterion-based multi-stage SAEA (EIC-MSSAEA) to tackle these challenges. Specifically, EIC-MSSAEA comprises three stages. In the first stage, we ignore constraints to facilitate the rapid traversal of infeasible obstacles. In the second stage, only one constraint is activated at a time to increase algorithm diversity. Finally, in the last stage, we activate all constraints to improve overall feasibility. In each stage, EIC-MSSAEA first employs NSGA-III as the underlying baseline solver to explore the search space, in which promising solutions are then selected by an ensemble infill criterion that incorporates multiple base-infill criteria to measure the feasibility, convergence, diversity, and uncertainty of candidate solutions. Experimental results demonstrate the competitiveness of EIC-MSSAEA against state-of-the-art SAEAs for ECMOPs.

Traditional stochastic configuration networks (SCNs)-based industrial soft sensors have the shortcomings of failing to account for “slowness” characteristic and struggling with processing rule-based information. The original slow feature extraction methods based on autoencoders with fixed structure are lack of flexibility and difficult to maintain a balance between efficiency and accuracy. To address these challenges, a framework of fuzzy hierarchical SCNs (FHSCNs) is proposed, which consists of a slow feature extraction block, a fuzzy inference block and an enhanced output block. The slow feature extraction block is designed, which utilizes a autoencoder based on two SCNs with shared-parameters and the incremental learning paradigm of SCNs to efficiently and adaptively extract the slow-varying latent features. The fuzzy inference block is proposed, which can process rule-based slow feature information. The fuzzy inference block can allow the model to have the fuzzy reasoning capabilities and improve the model interpretability. The enhanced output block with an enhancement layer and a direct-connect portion is presented, which enables the FHSCNs to have the ability of capturing both linearity and nonlinearity of the fuzzy rule-based features. The proposed framework is validated through comprehensive experiments to demonstrate its effectiveness in constructing industrial soft sensor model.

Adversarial attacks present a severe threat to the trustworthiness of remote sensing uncrewed systems. Existing detection methods are mostly limited to binary classification, lacking fine-grained traceability and relying heavily on adversarial example (AE) training. To overcome these challenges, we propose a training-free hierarchical adversarial attack traceability framework leveraging the zero-shot transfer ability of contrastive language-image pretraining (CLIP), eliminating the dependency on AE training. For the first time, we establish a multigranularity attack propagation path analysis system. Specifically, the framework leverages CLIP for fine-grained adversarial attack classification without fine-tuning, constructing a hierarchical traceability system (HTS) from coarse-grained to fine-grained semantic levels. Based on Bayesian inference, we design a hierarchical probability fusion method that improves coarse-grained results through maximum a posteriori (MAP) estimation of fine-grained classifiers, collaboratively optimizing hierarchical probability distributions. To capture frequency-domain characteristics of adversarial attacks, we propose a frequency-aware kernel approximation combining high-frequency-enhanced radial basis function (RBF) kernels with ridge regression, improving sensitivity to subtle perturbations in the reproducing kernel Hilbert space (RKHS). This study establishes a comprehensive traceability system for attack propagation chains, providing an interpretable, training-free paradigm for remote sensing security, significantly enhancing fine-grained detection capabilities of uncrewed systems. Extensive experiments on two public remote sensing datasets, RSSSN7 and DWEFS, and one proprietary dataset demonstrate that our method significantly improves adversarial classification accuracy across two victim models, achieving gains of 17.13%/18.89%/13.44% and 17.91%/19.58%/12.17%, respectively, compared with state-of-the-art (SOTA) training-free baselines.

In many real-world applications, the collected data with high dimensionality regularly also contain noise and outliers, which may result in learner model with insufficient generalization. The stochastic configuration network (SCN) is commonly applied to data analytics and achieve promising performance on industrial data modeling. However, the output weights of SCN are computed by the least-squares method, which is susceptible to noise and outliers. In this context, a novel framework of sparse generalized robust SCN (SGRSCN) is presented in this article. An arbitrary concave function is introduced in the loss function of SGRSCN that is employed to set the penalty weights for each training samples, which can reduce the negative impacts of noise and outliers to some extent. In addition, to address the architecture complexity and ill-posed problems of SCN when processing high-dimensional data, the loss function also consists of the${L}_{{2},{p}}$-norm ($\boldsymbol {0< }{p}\boldsymbol {< 1}$) regularization term. By adjustingp, the sparse model can be obtained. Then, the alternating optimization algorithm is applied to the optimization of loss function and the convergence analysis of the optimization process is provided. Moreover, a typical concave function is selected as an example using the framework of SGRSCN. Finally, the performance of SGRSCN is evaluated on two benchmark datasets and a real-world dataset with high-dimensional features. The experimental results demonstrate the superiority of SGRSCN.

Quality prediction is essential for optimizing operations and making timely decisions in industrial processes. However, the dynamic nature of industrial data, characterized by different sampling periods, presents significant challenges for quality prediction. The relationship between quality indices and various industrial data with differing sampling periods is complex and dynamically correlated in both spatial and temporal dimensions. To address this issue, we propose an adaptive multiresampling learning (AMRL) network that performs deep spatial-temporal feature mining and extraction from dual-scale data, facilitating multistep industrial quality prediction. The AMRL network leverages principal component scores of fast-scale process data and an adaptive multiresampling module to construct multiple fast-scale resampling channels adaptively. The cross-scale deep convolutional neural network and the multichannel self-attention module are then employed to capture spatial-temporal features and selectively focus on critical regions within the multiresampling sequences. We compared and evaluated the proposed method against eight state-of-the-art methods using real industrial datasets. The comparison results demonstrate the superior performance of the AMRL in multistep industrial quality prediction.

Class imbalance, which is common in real-world classification tasks, often leads to biased models favoring majority classes. Data oversampling is a widely used strategy to address this issue. However, traditional oversampling methods often generate incorrect or redundant instances when class overlap occurs, increasing decision boundary complexity. To this end, we propose a novel Generative Oversampling approach to addressing Class Imbalance and Overlap (GOIO) in the classification of tabular data. GOIO combines a Metric-Learning-based Variational Autoencoder (MLVAE) and a Conditional Latent Diffusion Model (CLDM) to handle class imbalance and overlap effectively. The MLVAE employs a triplet-center loss to the adverse effects of class overlap by transforming the data distribution into a more separable latent feature space. Following this, the CLDM is trained with class-center feature prompting and classifier-free guidance strategy to capture class-specific latent distributions accurately. Minority class samples are synthesized in the latent space using the CLDM and then reconstructed into the data space via the MLVAE decoder. Comprehensive experiments on 18 real-world and five synthetic datasets demonstrate that GOIO outperforms the state-of-the-art oversampling methods in F1-score, MCC, and Accuracy. Ablation studies further validate the effectiveness of the proposed contributions in addressing class imbalance and overlap.

Cross-project defect prediction (CPDP) leverages machine learning (ML) techniques to proactively identify software defects, especially where project-specific data is scarce. However, existing CPDP approaches suffer from three critical limitations: ineffective exploration of high-dimensional parameter spaces, poor adaptability across diverse projects with heterogeneous data distributions, and inadequate handling of feature redundancy and distribution discrepancies between source and target projects. To address these challenges, we formulate CPDP as a multi-objective bilevel optimization (MBLO) method, dubbed MBL-CPDP. Our approach comprises two nested problems: the upper-level, a multi-objective combinatorial optimization problem, enhances robustness by optimizing ML pipelines that integrate feature selection, transfer learning, and classification techniques, while the lower-level problem fine-tunes their hyperparameters. Unlike traditional methods that employ fragmented optimization strategies or single-objective approaches that introduce bias, MBL-CPDP provides a holistic, end-to-end optimization framework. Additionally, we propose an ensemble learning method to better capture cross-project distribution differences and improve generalization across diverse datasets. An MBLO algorithm is then presented to effectively solve the formulated MBLO problem. To evaluate MBL-CPDP’s performance, we compare it with five automated ML tools and 50 CPDP techniques across 20 projects. Extensive empirical results show that MBL-CPDP outperforms the comparison methods, demonstrating its superior adaptability and comprehensive performance evaluation capability.

Medium-scale expensive multiobjective optimization problems (EMOPs) present a significant challenge to most existing surrogate-assisted evolutionary algorithms (SAEAs). Because the algorithms must balance convergence and diversity with a limited number of fitness evaluations (FEs), while managing the uncertainty in surrogate predictions within the medium-scale decision space. Therefore, this work proposes a surrogate-assisted multiobjective evolutionary algorithm based on two-level model management (SAMOEA-TL2M) to effectively address medium-scale EMOPs. In SAMOEA-TL2M, infill solutions are selected using the proposed two-level model management strategy. In the first level, the estimated non-dominated solutions with good shift-based density estimation (SDE) values are selected for the balance of convergence and diversity. In the second level, the estimated non-dominated solutions and high uncertainty solutions are considered. To quantify uncertainty, an inverse distance weighting (IDW) is introduced. Moreover, an accuracy rate indicator (ARI) is proposed for the optimization state assessment, providing guidance for adaptively executing the two model management levels. Extensive experiments on three widely used instances and time-varying ratio error estimation (TREE) problems with up to 120 dimensions demonstrate the superiority of SAMOEA-TL2M over five state-of-the-art SAEAs in solving medium-scale EMOPs.

Robots typically encounter diverse tasks, bringing a significant challenge for motion control. Pretraining reinforcement learning (PRL) enables robots to adapt quickly to various tasks by exploiting reusable skills. The existing PRL methods often rely on datasets and human expert knowledge, struggle to discover diverse and dynamic skills, and exhibit generalization and adaptability to different types of robots and downstream tasks. This article proposes a novel PRL algorithm based on the central motor system mechanisms, which can discover diverse and dynamic skills without relying on data and expert knowledge, effectively enabling robots to tackle different types of downstream tasks. Inspired by the cerebellum’s role in balance control and skill storage within the central motor system, an intrinsic fused reward is introduced to explore dynamic skills and eliminate dependence on data and expert knowledge during pretraining. Drawing from the basal ganglia’s function in motor programming, a discrete skill encoding method is designed to increase the diversity of discovered skills, improving the performance of complex robots in challenging environments. Furthermore, incorporating the basal ganglia’s role in motor regulation, a skill activity function is proposed to generate skills at varying dynamic levels, thereby improving the adaptability of robots in multiple downstream tasks. The effectiveness of the proposed algorithm has been demonstrated through simulation experiments on four different morphological robots across multiple downstream tasks.

Surrogate-assisted evolutionary algorithms (SAEAs) are often used for expensive constrained multi-objective optimization problems (CMOPs). However, they exhibit poor convergence and diversity on complex CMOPs because they may consume more evaluations for approximating the constrained Pareto Front (CPF), i.e., the infeasible regions far from the CPF. To address this issue, we propose a surrogate-assisted coevolution-ary algorithm, which improves the convergence and diversity performance by searching on auxiliary problems together with the original CMOP. Firstly, we construct a surrogate model for each objective and constraint function. The primary population searches on both objective and constrained surrogate models, which approximate for the original constrained problem. The secondary population searches on an auxiliary problem, which is a multi-objective optimization problem (MOP) approximated by objective surrogate models. The two populations exchange information to enhance the convergence performance of the primary population. To improve the feasibility, the auxiliary problem is adopted with a certain probability. A novel infill criterion is proposed to guide the selection of high-quality solutions for updating the surrogate model, which relies on the feasible ratio of evaluated solutions. The performance of the proposed algorithm is evaluated on multi-objective and many-objective benchmarks to demonstrate its competitiveness.

Image recognition techniques have become the mainstream solution for indoor unmanned aerial vehicle (UAV) localization and navigation due to the absence of global positioning system. However, unlike autonomous vehicles that enjoy the dividends of the "big model" era, UAVs fail to deploy such models due to the hardware limitations. To this end, this paper proposes a compact all multilayer perceptron (MLP) deep neural network structure that offers a paradigm for theory-guided prompt structural compression of large-scale MLP models. First, we propose a gradient-based sensitivity analysis (GB-SA) method. Unlike existing SA methods, GB-SA obtains nodes’ sensitivity indices with gradient information, openning the possibility of efficient SA for large models. We start by integrating GB-SA with MLP and then extend the mode to MLP-Mixer, which is a promising all-MLP deep neural network. By deeply combining GB-SA and MLP-Mixer, SA-MLP-Mixer emerges as a compact model without reducing the model precision. Finally, we evaluate the effectiveness of the proposed model on the benchmark. The experimental results show that SA-MLP-Mixer has an airborne level model scale and accurate localization capability.

Refinery production scheduling is a mixed-integer programming problem, which exists the issue of combinational explosion. Thus, solving a large-scale refinery production scheduling problem is time-consuming. This article proposes an approximate solution framework based on reinforcement learning (RL) for large-scale long-time refinery production scheduling problems to rapidly obtain a satisfactory solution. In the proposed algorithm, the Proximal Policy Optimization algorithm is used to process the continuous action. To address the cold start issue of RL in refinery scheduling problem, we present an initialization method for the actor of agent, which utilizes the operation knowledge of tractable small-scale problems to initialize the actor network, and the agent is trained in the environment of large-scale problems. Hence, the convergence of the RL algorithm is greatly accelerated. In addition, the product flowrate concept is used to express the state, making the scheduling agent scalable in terms of scheduling horizon. Experimental studies show, to large-scale refinery scheduling problems, the proposed algorithm can obtain better solutions than that of the CPLEX solver and the existing evolutionary algorithm in a much shorter solving time of the two methods.Note to Practitioners—Scheduling is a link between planning and execution, and it can bring huge economic benefits to the refinery enterprises. With the enlargement of scheduling horizon, the scale of scheduling problems increases dramatically. How to deal with this large-scale scheduling problem caused by a long scheduling horizon is a significant problem. In this paper, the proposed method learns a decision-maker by reinforcement learning and applies to large-scale problems to obtain a good solution quickly. The proposed method is essentially a heuristic algorithm, and it is easy to implement in practice. At present, more and more things will be integrated into one model, leading to the traditional solver cannot meet the application needs. The fast solution method is necessary to be used to solve this problem in the new era.

Plate shape quality anomaly detection suffers from the problems of less labeled and multisource heterogeneous (MSH) data due to the complex production process with multi-process and multi-equipment. To address these problems, this paper proposes a self-supervised learning framework based on MSH contrastive learning (MSH-CL). In the proposed framework, an encoder is designed to extract the MSH information, which contains an MSH feature extract network (MSH-FEN) and an MSH feature fusion network (MSH-FFN). In MSH-FEN, a two-pathway CNN network is employed to extract the MSH feature. Then, a cross-MSH (CMSH) data fusion strategy based on the attention mechanism is designed to characterize the relationship between the MSH data in the MSH-FFN. The learning process consists of two phases, namely the self-supervised feature learning phase based on the contrastive learning (CL) and the supervised fine-tuning phase. The former phase is responsible for training the encoder with massive unlabeled data. Then, a classifier followed by the encoder is built with less labeled data in the supervised fine-tuning phase, which is applied to plate shape quality anomaly detection. Finally, the experiments are carried out on the real-world data set collected from a heavy-plate production process. The experimental results demonstrate the effectiveness and superiority of the proposed approach. Note to Practitioners—This paper focuses on the problem of shape quality anomaly detection for the heavy-plate production process, which is crucial to steel-making. However, the most quality-relevant data-driven modeling approaches suffer from the problem of requiring massive labeled data. In addition, the data generated in the practice heavy-plate production process have the characteristics of multisource heterogeneous (MSH). Motivated by the challenge of the less labeled and MSH data modeling, this paper proposed a self-supervised feature learning approach to learn a feature encoder to model the MSH data. Then, fine-tuning is used to obtain the model to detect the plate shape quality. In the feature encoder, a cross-MSH (CMSH) data fusion strategy is proposed. Finally, the effectiveness of the proposed approach is demonstrated by comparing it with other advanced algorithms. The proposed approach can be extended to other heavy-plate quality anomaly detection problems with less labeled and MSH data.

Alarm flood classification and alarm prediction are significant ways to assist the on-site operators to manage alarm floods and maintain process safety. The two tasks are interdependent considering the decisive role of different alarm floods on the arising alarms. To give comprehensive consideration of both, this work proposed a hierarchical strategy for alarm flood classification and alarm prediction leveraging the structure feature of alarm floods. The structure feature aims at revealing the sparse causal dependencies among alarm variables. It is achieved by a deep learning model under the guidance of a designed objective function that probabilistically parametrizes causal dependencies with sparsity constraint. Due to its interpretable physical meaning, desirable robustness and discriminate properties are achieved, allowing a win-win situation for both tasks. Based on the structure features, the hierarchical strategy is given, where an overall classifier is built while prediction models are trained for each category. The classifier trained by structure features is predisposed to generate satisfactory early classification results, enabling timely prediction. For the prediction, the structure features are also used to incorporate temporal features to achieve better performance. Experimental results illustrate the interpretability of structure features and show the feasibility of the proposed hierarchical strategy.Note to Practitioners—During alarm floods, the alarm patterns are different than usual and the generic prediction model may fail. The focus of this study is to achieve a win-win situation for both alarm flood classification and prediction, thereby providing comprehensive information required for handling alarm floods. Considering the arising alarm is strongly affected by the type of current alarm flood, a hierarchical alarm flood classification and alarm prediction strategy is given. It exploits the essential characteristic of alarm interactions in alarm floods to generate robust and early classification results, allowing timely predictions by category. In this way, the performance of both tasks can be guaranteed. The proposed method requires labeled historical alarm flood data.

Although reinforcement learning (RL) is effective in stabilizing systems, it faces many challenges in solving the tracking problem of unknown continuous-time systems. One of the major challenges is that RL-based control can hardly satisfy both the transient and steady-state performance requirements for the tracking problem simultaneously. In this study, instead of implementing an RL controller, the RL agent acts as a planner in the closed-loop reference model. The RL-based planner concentrates on tracking performance optimization by the constrained integral RL algorithm. Meanwhile, the system is controlled by the proposed library-based adaptive controller, which contains a library of candidate functions for modeling the unknown system dynamics. A natural gradient-like adaptive law is developed to update the controller, ensuring asymptotic tracking and promoting sparsity in the controller parameter. Compared with the conventional RL-based control, the proposed framework can eliminate the tracking error while avoiding the high-frequency oscillation and peaking phenomenon. Furthermore, we theoretically demonstrate that our approach can improve the transient performance in terms of the${\cal L}_{2} $norm of the tracking error and explicitly limit the${\cal L}_{\infty} $norm of the peaking value through the Lyapunov analysis. Simulations are presented to support the theoretical findings at the end of the paper.Note to Practitioners—Practical control design is often interested in tracking non-zero reference trajectories. However, the non-optimal transient response, such as oscillation and overshoot, is the major obstacle to the development of a high-performance tracking control system. The proposed method addresses this issue by designing a constrained RL-based CRM to ensure the optimal transient performance of the system. The primary advantage is that the maximum peaking value can be conveniently tuned as a hyperparameter by users, which is extremely useful in practice. Furthermore, the proposed library-based adaptive controller can handle unknown system dynamics, where the governing equations of the dynamics can be determined through engineering experience.

Bilevel optimization is a special type of optimization in which one problem is embedded within another. The bilevel optimization problem (BLOP) of which both levels are multiobjective functions is usually called the multiobjective BLOP (MBLOP). The expensive computation and nested features make it challenging to solve. Most existing studies look for complete lower-level solutions for every upper-level variable. However, not every lower-level solution will participate in the bilevel Pareto-optimal front. Under a limited computational budget, instead of wasting resources to find complete lower-level solutions that may not be in the feasible region or inducible region of the MBLOP, it is better to concentrate on finding the solutions with better performance. Bearing these considerations in mind, we propose a multiobjective bilevel optimization solving routine combined with a knee point driven algorithm. Specifically, the proposed algorithm aims to quickly find feasible solutions considering the lower-level constraints in the first stage and then concentrates the computational resources on finding solutions with better performance. Besides, we develop several multiobjective bilevel test problems with different properties, such as scalable, deceptive, convexity, and (dis)continuous. Finally, the performance of the algorithm is validated on a practical petroleum refining bilevel problem, which involves a multiobjective environmental regulation problem and a petroleum refining operational problem. Comprehensive experiments fully demonstrate the effectiveness of our presented algorithm in solving MBLOPs.

Dynamic environments pose great challenges for expensive optimization problems, as the objective functions of these problems change over time and thus require remarkable computational resources to track the optimal solutions. Although data-driven evolutionary optimization and Bayesian optimization (BO) approaches have shown promise in solving expensive optimization problems in static environments, the attempts to develop such approaches in dynamic environments remain rarely explored. In this article, we propose a simple yet effective meta-learning-based optimization framework for solving the expensive dynamic optimization problems. This framework is flexible, allowing any off-the-shelf continuously differentiable surrogate model to be used in a plug-in manner, either in data-driven evolutionary optimization or BO approaches. In particular, the framework consists of two unique components: 1) the meta-learning component, in which a gradient-based meta-learning approach is adopted to learn experience (effective model parameters) across different dynamics along the optimization process and 2) the adaptation component, where the learned experience (model parameters) is used as the initial parameters for fast adaptation in the dynamic environment based on few shot samples. By doing so, the optimization process is able to quickly initiate the search in a new environment within a strictly restricted computational budget. Experiments demonstrate the effectiveness of the proposed algorithm framework compared to several state-of-the-art algorithms on common benchmark test problems under different dynamic characteristics.

Efficient monitoring of production performance is crucial for ensuring safe operations and enhancing the economic benefits of the Iron and Steel Corporation. Although basic modeling algorithms and visualization diagrams are available in many scientific platforms and industrial applications, there is still a lack of customized research in production performance monitoring. Therefore, this article proposes an interactive visual analytics approach for monitoring the heavy-plate production process (iHPPPVis). Specifically, a multicategory aggregated monitoring framework is proposed to facilitate production performance monitoring under varying working conditions. In addition, A set of visualizations and interactions are designed to enhance analysts' analysis, identification, and perception of the abnormal production performance in heavy-plate production data. Ultimately, the efficacy and practicality of iHPPPVis are demonstrated through multiple evaluations.

Dynamic constrained multiobjective optimization involves irregular changes in the distribution of the true Pareto-optimal fronts, drastic changes in the feasible region caused by constraints, and the movement directions and magnitudes of the optimal distance variables due to diverse changing environments. To solve these problems, we propose a multi-population evolution based dynamic constrained multiobjective optimization algorithm. In this algorithm, we design a tribe classification operator to divide the population into different tribes according to a feasibility check and the objective values, which is beneficial for driving the population toward the feasible region and Pareto-optimal fronts. Meanwhile, a population selection strategy is proposed to identify promising solutions from tribes and exploit them to update the population. The optimal values of the distance variables vary differently with dynamic environments, thus, we design a dynamic response strategy for solutions in different tribes that estimates their distances to approach the Pareto-optimal fronts and regenerates a promising population when detecting environmental changes. In addition, a scalable generator is designed to simulate diverse movement directions and magnitudes of the optimal distance variables in real-world problems under dynamic environments, obtaining a set of improved test problems. Experimental results show the effectiveness of test problems, and the proposed algorithm is impressively competitive with several chosen state-of-the-art competitors.

Evolutionary multitasking Optimization (EMTO) is a paradigm that optimizes multiple tasks simultaneously to improve the overall performance of all tasks by seamlessly transferring useful knowledge among them. Although EMTO has received significant interest, rare studies consider handling tasks that are multimodal optimization problems (MMOPs) with multiple global optimal solutions. Due to the multiple different modalities of each task, a major challenge of solving multiple MMOPs is how to extract and transfer knowledge across modalities of different tasks. To this end, this paper designs a distributed knowledge transfer based evolutionary multitask multimodal optimization (EMTMO-DKT) approach for solving multiple MMOPs simultaneously by discovering and utilizing local knowledge across modalities of different tasks. Specifically, we first divide the population of each task into multiple subpopulations, where each subpopulation explores a modality. Then, we propose an evolution path based similarity measurement to measure the local similarities between subpopulations of different tasks. Since the modalities can be locally similar across tasks, we develop a subpopulation cross matching strategy according to the obtained similarities to pair subpopulations of different tasks. In this stage, the successfully paired subpopulations are allowed to transfer knowledge. Finally, the knowledge transfer probability self-adjusting strategy is applied to each subpopulation to balance knowledge transfer and self-evolution, so as to improve search efficiency. In this paper, a set of multitask multimodal optimization test problems are constructed to assess the efficacy of compared algorithms. Experimental results on both the benchmark functions and the real-world optimization problem demonstrate that the proposed algorithm can quickly locate more global optima in comparison with state-of-the-art EMTO and multimodal optimization algorithms.

In some real-world applications, the optimization problems may involve multiple design stages. At each design stage, the objective is incrementally modified by incorporating more decision variables and optimized. In addition, the fitness evaluations (FEs) are often highly costly. Such optimization problems can be called expensive incremental optimization problems (EIOPs). Despite their importance, EIOPs have not attracted much attention over the past few years. Since the objectives of different design stages are different but related, reusing the search experience from the past design stages is beneficial to the evolutionary search of the current design stage. Therefore, a surrogate-assisted differential evolution with knowledge transfer (SADE-KT) is proposed in this work, which aims to fill the current gap in solving EIOPs. The major merit of the proposed SADE-KT is its ability to seamlessly integrate knowledge transfer and the surrogate-assisted evolutionary search. In SADE-KT, a surrogate based hybrid knowledge transfer strategy is first proposed. This strategy makes it possible to reuse the knowledge captured from the past design stages by leveraging different knowledge transfer techniques. As a result, the convergence for the current design stage can be speeded up. Then, a two-level surrogate-assisted evolutionary search is developed to search for the optimum. Comprehensive empirical studies have demonstrated that the proposed algorithm works efficiently on EIOPs.

Zero-shot fault diagnosis can identify unseen faults by predicting attributes. However, existing methods ignore the multi-grained characteristics of attributes, namely the varying levels of detail in describing fault categories. We recognize the following considerations for the first time: (1) attributes show typical multi-grained characteristics, which could be expressed in a coarse-to-fine-grained hierarchical structure; (2) multi-grained attributes play different roles in fault diagnosis, where coarse-grained attributes indicate the rough range of faults, while fine-grained attributes facilitate the precise identification of fault types. In this paper, a fuzzy hierarchical zero-shot learning method is proposed to solve these issues. First, the attributes are divided into different layers according to the coarse-to-fine granularity via expert knowledge rather than being treated equally. Then, a knowledge transfer strategy is designed to transfer the knowledge from coarse-grained attributes to fine-grained ones, which can improve attribute prediction accuracy. Finally, a fuzzy inference strategy is developed to distinguish the effect of attributes with different granularity on fault inference. This strategy can identify the faults stepwise in a coarse-to-fine-grained order. The effectiveness of the proposed method is verified by a real thermal power plant process.

Modern industrial processes are in general characterized by multiple operation modes in response to diverse product requirements, and it is challenging to monitor such complex processes. Most of the existing multimodal monitoring strategies tend to model each mode individually, where relationships among different modes are ignored, leaving room for improving process monitoring performance. In this article, we propose a new monitoring approach called low-rank multimanifold embedding (LRME) for monitoring complex industrial processes with multiple operation modes, in which individualities in each mode, connections between modes, and commonalities among all modes are captured simultaneously. LRME can learn the lowest rank representation shared by all modes, revealing the common information among all modes. The within-mode and between-mode adjacency matrices are constituted to characterize the within-mode individualities and between-mode connections, which enables the reduced-dimensional representations extracted faithfully to reveal more intrinsic characterizations of multimodal data and further enhance the process monitoring capability. In addition, a point-to-manifold metric is developed to identify the running-on operation mode during online monitoring. The monitoring results on the Tennessee Eastman process and a practical industrial process demonstrate the effectiveness of our proposed approach.

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.

Process shift of multimode process involving data distribution and dynamic relation makes traditional transfer learning methods be intractable and even result in negative transfer. To tackle this issue, this article proposes a novel self-tuning transfer dynamic modeling method for quality prediction of multimode processes. First, in order to capture domain-invariant spatiotemporal (DIST) features, a transfer dynamic convolution autoencoder (TDCAE) with a feature decomposition structure is established. Meanwhile, a first-order vector autoregressive constraint is embedded to extract consistent inner dynamics for DIST features. Then, a shared regression network is established to extract the relations with quality variables. Furthermore, by making full use of private spatiotemporal information from target labeled samples in response to the process shift, the self-tuning TDCAE (STDCAE) aided by a fine-tuning strategy is established for online compensation. Finally, the efficacy of the proposed TDCAE and STDCAE is demonstrated by a comprehensive study of a three-phase flow facility process.

Modern industrial process modeling requires models to adapt quickly to real-time operating conditions. To this end, this article proposes a fast sensitivity analysis (SA)-based self-organizing broad learning system (SASO-BLS) that offers a paradigm for theory-guided online structural self-adaptation of differentiable models. Specifically, SASO-BLS is implemented by embedding in BLS an interpretable and efficient model compression method called fast partial differential-based SA (FPD-SA). Unlike conventional SA methods that require iterative evaluation of SA indexes for all samples, FPD-SA exploits the deduced chain rule across categories, effectively mitigating the computational burden imposed by industrial Big Data and reducing computational time. In addition, we derive the offline SASO-BLS algorithm for discrete data and extend it to the online scenario for real-time streaming data processing. Note that both modes obviate the necessity of recalculating the pseudo-inverse of the entire state matrix, facilitating SASO-BLS in attaining remarkable efficiency in structural self-organization. Finally, a theoretical justification of the universal approximation property for SASO-BLS is presented. Experimental results on a fault diagnosis benchmark dataset and a real industrial process one demonstrate the effectiveness of the proposed approach.

In this article, a model-free Q-learning algorithm is proposed to solve the tracking problem of linear discrete-time systems with completely unknown system dynamics. To eliminate tracking errors, a performance index of the Q-learning approach is formulated, which can transform the tracking problem into a regulation one. Compared with the existing adaptive dynamic programming (ADP) methods and Q-learning approaches, the proposed performance index adds a product term composed of a gain matrix and the reference tracking trajectory to the control input quadratic form. In addition, without requiring any prior knowledge of the dynamics of the original controlled system and command generator, the control policy obtained by the proposed approach can be deduced by an iterative technique relying on the online information of the system state, the control input, and the reference tracking trajectory. In each iteration of the proposed method, the desired control input can be updated by the iterative criteria derived from a precondition of the controlled system and the reference tracking trajectory, which ensures that the obtained control policy can eliminate tracking errors in theory. Moreover, to effectively use less data to obtain the optimal control policy, the off-policy approach is introduced into the proposed algorithm. Finally, the effectiveness of the proposed algorithm is verified by a numerical simulation.

Typically, industrial processes possess both temporal and spatial dependencies due to intravariable dynamics and intervariable couplings. The two dependencies have different manifestations, indicating diverse process characteristics. However, the existing methods fail to separate temporal and spatial information well, leading to inappropriate representation and inaccurate fault detection and isolation results. This study proposes an explicit representation and customized fault isolation framework to tackle temporal and spatial characteristics, so as to identify and locate anomalies affecting different dependencies. First, we design a double-level separation method for temporal and spatial information. In the first level, we construct two independent auto-encoding modules to extract temporal correlation and spatial graph structure in parallel. In the second level, we propose an information aliasing loss function to guild the two modules to distinguish between temporal and spatial characteristics, further facilitating information separation. By monitoring the explicit temporal and spatial statistics obtained by the two modules, spatiotemporal dependencies of anomalies can be determined for subsequent isolation. Furthermore, we propose a customized isolation strategy for anomalies in temporal and spatial characteristics. By quantifying changes in intravariable temporal dynamics and intervariable spatial graph structure individually, temporal impact and spatial propagation of faults can be finely characterized and isolated. Three examples are adopted to verify the performance of the proposed framework, including a numerical example, a real condensing system of the thermal power plant process, and the Tennessee Eastman benchmark process.

Abstract Data-driven fault detection techniques have attracted extensive attention in engineering, industry and many other areas in recent years. In many real applications, the following situation often occurs: data for certain types of faults (unseen faults) are not available to train models that are used for fault detection. Such a scenario can occur when data collection becomes highly time-consuming or destructive. To address this challenging problem, a novel fault detection method using zero-shot learning (ZSL) is proposed in this paper, which contains three phases: feature extraction, label embedding, and feature embedding. The method first extracts features from raw signals by applying a one-dimensional convolutional neural network (1D CNN), then builds semantic descriptions (human-defined) as fault attributes shared between seen faults and unseen faults, and finally uses a bi-linear compatibility function to find the highest-ranking fault type. The proposed semantic space based zero-shot learning with 1D CNN is called SSB-ZSL-1DCNN. The cosine distance is used to measure the similarity between feature embeddings and fault attributes. An important characteristic of SSB-ZSL-1DCNN is that the model, trained using only samples of seen faults, can be used to detect unseen defects. To evaluate the proposed method, two case studies are designed based on two well-known benchmarks (the Tennessee-Eastman chemical control process and the rolling bearing experiments at the Case Western Reserve University, respectively). The results demonstrate that the proposed method shows remarkable performance in detecting unseen faults.

Existing work on offline data-driven optimization mainly focuses on problems in static environments, and little attention has been paid to problems in dynamic environments. Offline data-driven optimization in dynamic environments is a challenging problem because the distribution of collected data varies over time, requiring surrogate models and optimal solutions tracking with time. This paper proposes a knowledge-transfer-based data-driven optimization algorithm to address these issues. First, an ensemble learning method is adopted to train surrogate models to leverage the knowledge of data in historical environments as well as adapt to new environments. Specifically, given data in a new environment, a model is constructed with the new data, and the preserved models of historical environments are further trained with the new data. Then, these models are considered to be base learners and combined as an ensemble surrogate model. After that, all base learners and the ensemble surrogate model are simultaneously optimized in a multitask environment for finding optimal solutions for real fitness functions. In this way, the optimization tasks in the previous environments can be used to accelerate the tracking of the optimum in the current environment. Since the ensemble model is the most accurate surrogate, we assign more individuals to the ensemble surrogate than its base learners. Empirical results on six dynamic optimization benchmark problems demonstrate the effectiveness of the proposed algorithm compared with four state-of-the-art offline data-driven optimization algorithms. Code is available at https://github.com/Peacefulyang/DSE_MFS.git.

This article investigates the sensitivity analysis (SA) of high-dimensional data to identify the effects of process variables on output quantity of interest (QoI) in industrial soft sensor modeling. The computational cost of analyzing the SA of high-dimensional data is high, and models available for SA techniques usually have limited generalization capacity. Therefore, we propose a novel high-dimensional data global SA (GSA) approach based on a deep soft sensor model to address these issues. We first develop an approximately incremental grouping (AIG) algorithm and a region-based cooperative co-evolution (RBCC) algorithm to decompose the high-dimensional data into independent regions for the GSA. Subsequently, a multihead deep soft sensor model with generalization performance is designed to determine the GSA indices of each decomposed region. Specifically, the region of interest (RoI) align algorithm provides the multihead with precisely located decomposed region features. Finally, based on the uncertainty analysis of each model head, we present a joint loss function with the Monte Carlo dropout (MC-dropout) algorithm to measure the GSA indices of each decomposed region on QoIs. Experimental evaluation results on a benchmark dataset and a real-world one demonstrate the effectiveness of the proposed approach in addressing the GSA of high-dimensional data in industrial processes.

Operational data from industrial processes typically consist of continuous and categorical variables. They reveal different aspects of the operational conditions, and both are useful for comprehensive fault detection and diagnosis (FDD). However, due to multiple production modes, the variables usually do not follow Gaussian or Bernoulli distributions. Furthermore, their correlations can be different across normal and various faulty classes. Thus, the main challenge is how to fuse the complementary information in the two types of variables and accurately characterize the complicated distribution for each class. This article proposes a flexible probabilistic framework that can concurrently analyze continuous and categorical variables for FDD. Our framework specifies a finite mixture model for each class. Thus, it can handle non-Gaussian and non-Bernoulli variables and capture their correlations under the conditional independence assumption. We then introduce the variational inference for parameter estimation, which makes our framework adaptive to the different distributions of various classes. Furthermore, a unified statistical index is designed, which gives our method extra capability to distinguish unknown faults. Finally, the effectiveness of our method is validated on the Tennessee Eastman (TE) process and a practical industrial plant process. The averageF1score of our method is improved by 3.4/3.9 percentage points in the single-mode/multimode situation on the TE process compared with traditional mixture discriminant analysis. In the industrial plant process, when unknown faults are added, the averageF1score of our method is 91.1% and only 1.2 percentage points lower than that on the test set without unknown faults.

This work addresses unsupervised partial domain adaptation (PDA), in which classes in the target domain are a subset of the source domain. The key challenges of PDA are how to leverage source samples in the shared classes to promote positive transfer and filter out the irrelevant source samples to mitigate negative transfer. Existing PDA methods based on adversarial DA do not consider the loss of class discriminative representation. To this end, this article proposes a contrastive learning-assisted alignment (CLA) approach for PDA to jointly align distributions across domains for better adaptation and to reweight source instances to reduce the contribution of outlier instances. A contrastive learning-assisted conditional alignment (CLCA) strategy is presented for distribution alignment. CLCA first exploits contrastive losses to discover the class discriminative information in both domains. It then employs a contrastive loss to match the clusters across the two domains based on adversarial domain learning. In this respect, CLCA attempts to reduce the domain discrepancy by matching the class-conditional and marginal distributions. Moreover, a new reweighting scheme is developed to improve the quality of weights estimation, which explores information from both the source and the target domains. Empirical results on several benchmark datasets demonstrate that the proposed CLA outperforms the existing state-of-the-art PDA methods.

This article studies an operational optimization problem of the fluid catalytic cracking (FCC) unit under uncertainty. The objective of this problem is to quickly reoptimize the operating variables that control the operational condition of the FCC unit when fossil fuel yield constraints or prices change. To solve this problem, based on the challenges caused by the varied constraints, we establish a mathematical model and propose a fast adaptive differential evolution algorithm with an adaptive mutation strategy, a parameter adaptation strategy, a repaired strategy, and an enhanced strategy. In the proposed algorithm, we integrate the status information of each solution into the mutation strategy and parameter adaptation scheme to search for the best solution in the irregular feasible region of the operating variables. In addition, a repaired strategy is proposed to repair the infeasible operating variables with unknown bounds, and an enhanced strategy is presented to further improve the objective function value of the best solution. The experimental results on ten test scenarios with different fossil fuel yield constraints and prices demonstrate the robustness of the proposed algorithm for optimizing the operating variables of the FCC unit under uncertainty.

For dynamic multiobjective optimization problems (DMOPs), it is challenging to track the varying Pareto-optimal front. Most traditional approaches estimate the Pareto-optimal sets in the decision space. However, the obtained solutions do not necessarily satisfy the desired properties of decision makers in the objective space. Inverse model-based algorithms have a great potential to solve such problems. Nonetheless, the existing ones have low precision for handling DMOPs with nonlinear correlations between the objective and decision vectors, which greatly limits the application of the inverse models. In this article, an inverse Gaussian process (IGP)-based prediction approach for solving DMOPs is proposed. Unlike most traditional approaches, this approach exploits the IGP to construct a predictor that maps the historical optimal solutions from the objective space to the decision space. A sampling mechanism is developed for generating sample points in the objective space. Then, the IGP-based predictor is employed to generate an effective initial population by using these sample points. The proposed method by introducing IGP can obtain solutions with better diversity and convergence in the objective space, which is more responsive to the demand of decision makers than the traditional methods. It also has better performance than other inverse model-based methods in solving nonlinear DMOPs. To investigate the performance of the proposed approach, experiments have been conducted on 23 benchmark problems and a real-world raw ore allocation problem in mineral processing. The experimental results demonstrate that the proposed algorithm can significantly improve the dynamic optimization performance and has certain practical significance for solving real-world DMOPs.

Decomposition algorithms for large-scale refinery scheduling problems commonly adopt the spatially dividing method. When the scheduling horizon further enlarges, the sizes of the resulting subproblems exponentially grow, leading to the performance degradation of the decomposition algorithms. In this article, to solve this issue about subproblem size caused by the long-time scheduling horizon, a knowledge transfer based algorithm is proposed, where the size of subproblems is constant and a novel concept of product flowrate is introduced to transfer operation knowledge from tractable small-scale short-time problems to intractable large-scale long-time problems to quickly obtain a satisfactory solution for large-scale problems. Experimental results show that the proposed algorithm can achieve better solutions within 10−44 s for large-scale refinery scheduling problems with 50−200 time slots (the existing model is with less than 30 time slots), compared to the CPLEX solver with a running time of 3600 s and the existing evolutionary algorithm with a running time of 100−500 s. The results also show that, at present, the heuristic method (the essence of the proposed algorithm) is the main tool to solve large-scale combinatorial optimization problems, and even in the CPLEX solver, the heuristic is also the most active part.

This article investigates adaptive robust controller design for discrete-time (DT) affine nonlinear systems using an adaptive dynamic programming. A novel adaptive interleaved reinforcement learning algorithm is developed for finding a robust controller of DT affine nonlinear systems subject to matched or unmatched uncertainties. To this end, the robust control problem is converted into the optimal control problem for nominal systems by selecting an appropriate utility function. The performance evaluation and control policy update combined with neural networks approximation are alternately implemented at each time step for solving a simplified Hamilton-Jacobi-Bellman (HJB) equation such that the uniformly ultimately bounded (UUB) stability of DT affine nonlinear systems can be guaranteed, allowing for all realization of unknown bounded uncertainties. The rigorously theoretical proofs of convergence of the proposed interleaved RL algorithm and UUB stability of uncertain systems are provided. Simulation results are given to verify the effectiveness of the proposed method.

To address the architecture complexity and ill-posed problems of neural networks when dealing with high-dimensional data, this article presents a Bayesian-learning-based sparse stochastic configuration network (SCN) (BSSCN). The BSSCN inherits the basic idea of training an SCN in the Bayesian framework but replaces the common Gaussian distribution with a Laplace one as the prior distribution of the output weights of SCN. Meanwhile, a lower bound of the Laplace sparse prior distribution using a two-level hierarchical prior is adopted based on which an approximate Gaussian posterior with sparse property is obtained. It leads to the facilitation of training the BSSCN, and the analytical solution for output weights of BSSCN can be obtained. Furthermore, the hyperparameter estimation process is derived by maximizing the corresponding lower bound of the marginal likelihood function based on the expectation-maximization algorithm. In addition, considering the uncertainties caused by both noises in the real-world data and model mismatch, a bootstrap ensemble strategy using BSSCN is designed to construct the prediction intervals (PIs) of the target variables. The experimental results on three benchmark data sets and two real-world high-dimensional data sets demonstrate the effectiveness of the proposed method in terms of both prediction accuracy and quality of the constructed PIs.

Existing transfer learning methods that focus on problems in stationary environments are not usually applicable to dynamic environments, where concept drift may occur. To the best of our knowledge, the concept drift-tolerant transfer learning (CDTL), whose major challenge is the need to adapt the target model and knowledge of source domains to the changing environments, has yet to be well explored in the literature. This article, therefore, proposes a hybrid ensemble approach to deal with the CDTL problem provided that data in the target domain are generated in a streaming chunk-by-chunk manner from nonstationary environments. At each time step, a class-wise weighted ensemble is presented to adapt the model of target domains to new environments. It assigns a weight vector for each classifier generated from the previous data chunks to allow each class of the current data leveraging historical knowledge independently. Then, a domain-wise weighted ensemble is introduced to combine the source and target models to select useful knowledge of each domain. The source models are updated with the source instances performed by the proposed adaptive weighted CORrelation ALignment (AW-CORAL). AW-CORAL iteratively minimizes domain discrepancy meanwhile decreases the effect of unrelated source instances. In this way, positive knowledge of source domains can be potentially promoted while negative knowledge is reduced. Empirical studies on synthetic and real benchmark data sets demonstrate the effectiveness of the proposed algorithm.

Gaussian processes (GPs) are widely used in surrogate-assisted evolutionary optimization of expensive problems mainly due to the ability to provide a confidence level of their outputs, making it possible to adopt principled surrogate management methods, such as the acquisition function used in the Bayesian optimization. Unfortunately, GPs become less practical for high-dimensional multiobjective and many-objective optimization as their computational complexity is cubic in the number of training samples. In this article, we propose a computationally efficient dropout neural network (EDN) to replace the Gaussian process and a new model management strategy to achieve a good balance between convergence and diversity for assisting evolutionary algorithms to solve high-dimensional multiobjective and many-objective expensive optimization problems. While the conventional dropout neural network needs to save a large number of network models during the training for calculating the confidence level, only one single network model is needed in the EDN to estimate the fitness and its confidence level by randomly ignoring neurons in both training and testing the neural network. Extensive experimental studies on benchmark problems with up to 100 decision variables and 20 objectives demonstrate that, compared to state of the art, the proposed algorithm is not only highly competitive in performance but also computationally more scalable to high-dimensional many-objective optimization problems. Finally, the proposed algorithm is validated on an operational optimization problem of crude oil distillation units, further confirming its capability of handling expensive problems given a limited computational budget.

As an essential component in multi- and many-objective optimization, decision-making process either selects a subset of solutions from the whole Pareto front or guides the search toward a small part of the Pareto front during the evolutionary process. In recent years, for many-objective optimization problems (MaOPs), a number of evolutionary algorithms have been developed to search for Pareto optimal solutions. However, there is a lack of research works focusing on designing decision-making approaches. In order to overcome this deficiency, we propose a novel knee-based decision-making method to search for several solutions of interest (SOIs) from a large number of solutions on the Pareto front, each of which contains the best convergence performance at least within its neighborhood and can be identified as a global or local knee solution. The optimization performance achieved by all SOIs approximates the performance of the whole Pareto front as much as possible. Furthermore, in order to relieve the difficulties in the decision-making process on MaOPs, a new visualization approach is developed based on this proposed decision-making approach. It provides information about the shape and location of the Pareto front, the possible bulge, as well as the convergence degree and distribution of solutions. The experimental results on several benchmark functions demonstrate the superiority of the proposed design in the selection of SOIs and visualization of high-dimensional objective space.

This Special Section presents the latest developments on intelligent modeling, neural networks, deep learning, and adaptive estimation, and their applications in industrial applications. Through a rigorous peer-review process, eleven articles have been accepted, which are summarized below.

The plant-wide production process is composed of multiple unit processes, in which the operational indices of each unit process are assigned and adjusted according to product quality, yield, and actual operating modes. Due to the changing operational conditions of the production process, the operational indices cannot be effectively adjusted by most of the model-based methods or evolutionary computation. In this article, the decision making of operational indices is formulated as a continuous state, continuous action reinforcement learning (RL) problem and a model-free RL algorithm is proposed, which learns a decision policy to determine the operational indices according to the actual operational conditions. Different from the existing methods, this article presents a multiactor networks ensemble algorithm and an actor-critic framework with stochastic policy to avoid falling into local optimums. The relatively overall optimal policy is obtained by extracting the results of parallel training of multiactor networks, which guarantees the optimality of the obtained policy. In addition, by using the experience replay, it is particularly valuable to effectively deal with the problem that lacking of sampling data in the model-free RL. Simulation studies are conducted on actual data of a mineral processing plant and the results demonstrate the effectiveness of the proposed algorithm.

Complex industrial processes are commonly characterized with multiple operation modes. The existing manifold learning-based process monitoring methods describe each mode individually without capturing the connections among different modes, which may deteriorate the monitoring capability. This article proposes a novel dimensionality reduction model referred as to multimanifold joint projections to monitor the multimode processes, where the intramode and the intermode adjacency matrices are constructed to reflect the underlying features within each mode and among different modes, respectively. The neighboring and nonneighboring structures of data within each mode are captured by the distance and angle information of pairwise points to reveal the intrinsic structure of the original data, thus offering a more faithful representation of multimodal data and further enhancing monitoring performance. During online monitoring, a point to manifold distance criterion is proposed to determine the running-on mode of new samples. Two case studies demonstrated the superior performance of the proposed approach in multimode process monitoring.

To achieve plant-wide operational optimization and dynamic adjustment of operational index for an industrial process, knowledge-based methods have been widely employed over the past years. However, the extraction of knowledge base is a bottleneck for most existing approaches. To address this problem, we propose a novel framework based on the generative adversarial networks (GANs), termed as decision-making GAN (DMGAN), which directly learns from operational data and performs human-level decision making of the operational indices for plant-wide operation. In the proposed DMGAN, two adversarial criteria and three cycle consistency criteria are incorporated to encourage efficient posterior inference. To improve the generalization power of a generator with an increasing complexity of the industrial processes, a reinforced U-Net (RU-Net) is presented that improves the traditional U-Net by providing a more general combinator, a building block design, and drop-level regularization. In this article, we also propose three quantitative metrics for assessing the plant-wide operation performance. A case study based on the largest mineral processing factory in Western China is carried out, and the experimental results demonstrate the promising performance of the proposed DMGAN when compared with decision-making based on domain experts.

This article presents a novel technique to achieve plant-wide performance optimization for large-scale unknown industrial processes by integrating the reinforcement learning method with the multiagent game theory. A main advantage of this technique is that plant-wide optimal performance is achieved by a distributed approach where multiple agents solve simplified local nonzero-sum optimization problems so that a global Nash equilibrium is reached. To this end, first, the plant-wide performance optimization problem is reformulated by decomposition into local optimization subproblems for each production index in a multiagent framework. Then, the nonzero-sum graphical game theory is utilized to compute the operational indices for each unit process with the purpose of reaching the global Nash equilibrium, resulting in production indices following their prescribed target values. The stability and the global Nash equilibrium of this multiagent graphical game solution are rigorously proved. The reinforcement learning methods are then developed for each agent to solve the nonzero-sum graphical game problem using data measurements available in the system in real time. The plant dynamics do not have to be known. Finally, the emulation results are given to show the effectiveness of the proposed automated decision algorithm by using measured data from a large mineral processing plant in Gansu Province, China.

To promote research on dynamic constrained multiobjective optimization, we first propose a group of generic test problems with challenging characteristics, including different modes of the true Pareto front (e.g., convexity-concavity and connectedness-disconnectedness) and the changing feasible region. Subsequently, motivated by the challenges presented by dynamism and constraints, we design a dynamic constrained multiobjective optimization algorithm with a nondominated solution selection operator, a mating selection strategy, a population selection operator, a change detection method, and a change response strategy. The designed nondominated solution selection operator can obtain a nondominated population with diversity when the environment changes. The mating selection strategy and population selection operator can adaptively handle infeasible solutions. If a change is detected, the proposed change response strategy reuses some portion of the old solutions in combination with randomly generated solutions to reinitialize the population, and a steady-state update method is designed to improve the retained previous solutions. The experimental results show that the proposed test problems can be used to clearly distinguish the performance of algorithms, and that the proposed algorithm is very competitive for solving dynamic constrained multiobjective optimization problems in comparison with state-of-the-art algorithms.

In offline data-driven optimization, only historical data is available for optimization, making it impossible to validate the obtained solutions during the optimization. To address these difficulties, this paper proposes an evolutionary algorithm assisted by two surrogates, one coarse model and one fine model. The coarse surrogate (CS) aims to guide the algorithm to quickly find a promising subregion in the search space, whereas the fine one focuses on leveraging good solutions according to the knowledge transferred from the CS. Since the obtained Pareto optimal solutions have not been validated using the real fitness function, a technique for generating the final optimal solutions is suggested. All achieved solutions during the whole optimization process are grouped into a number of clusters according to a set of reference vectors. Then, the solutions in each cluster are averaged and outputted as the final solution of that cluster. The proposed algorithm is compared with its three variants and two state-of-the-art offline data-driven multiobjective algorithms on eight benchmark problems to demonstrate its effectiveness. Finally, the proposed algorithm is successfully applied to an operational indices optimization problem in beneficiation processes.

It is challenging to develop point prediction models with high accuracy due to that outliers and noise are commonly present in the real-world data. In this context, this article proposes a novel robust stochastic configuration network (SCN) and uses the bootstrap ensemble strategy to construct prediction intervals (PIs). Since the output weights of the original SCN are computed by the least-squares method, which is sensitive to noise with an unknown distribution or outliers, a robust SCN based on a mixture of the Gaussian and Laplace distributions (MoGL-SCN) in the Bayesian framework is proposed. The mixed distributions can effectively characterize the complex distributions of the real-world data, and their heavy-tailed properties can improve the robustness of SCNs. Furthermore, there are no analytical solutions available to obtain the network parameters due to the assumption on the mixed distributions, hence, the parameters of the MoGL-SCN are estimated by the expectation-maximization algorithm. In addition, considering the uncertainties caused by both the model mismatch and noise in the real-world data, a bootstrap ensemble strategy using MoGL-SCN is designed to construct the PIs. The experimental results on two benchmark datasets and a real-world dataset demonstrate the effectiveness of the proposed method in terms of the quality of PIs, prediction accuracy, and robustness.

Obtaining accurate point prediction of industrial processes' key variables is challenging due to the outliers and noise that are common in industrial data. Hence the prediction intervals (PIs) have been widely adopted to quantify the uncertainty related to the point prediction. In order to improve the prediction accuracy and quantify the level of uncertainty associated with the point prediction, this article estimates the PIs by using ensemble stochastic configuration networks (SCNs) and bootstrap method. The estimated PIs can guarantee both the modeling stability and computational efficiency. To encourage the cooperation among the base SCNs and improve the robustness of the ensemble SCNs when the training data are contaminated with noise and outliers, a simultaneous robust training method of the ensemble SCNs is developed based on the Bayesian ridge regression and M-estimate. Moreover, the hyperparameters of the assumed distributions over noise and output weights of the ensemble SCNs are estimated by the expectation-maximization (EM) algorithm, which can result in the optimal PIs and better prediction accuracy. Finally, the performance of the proposed approach is evaluated on three benchmark data sets and a real-world data set collected from a refinery. The experimental results demonstrate that the proposed approach exhibits better performance in terms of the quality of PIs, prediction accuracy, and robustness.

This paper proposes a reference vector based multidirectional prediction strategy to address dynamic multiobjective optimization problems (DMOPs), whose PSs rotate with time. In this strategy, several reference vectors partition the population into multiple clusters and the same prediction trajectory is adapted in a cluster. In addition, a new offspring creation strategy in genetic operator named reference vector based offspring creation is proposed to accelerate the convergence and maintain the diversity. Experiments on seven benchmark problems are carried out to examine the performance of the proposed algorithm and the statistical results shows the proposed algorithm can address DMOPs with rotating PSs.

This paper studies a hot rolling batch scheduling problem in compact strip production (CSP), which is decomposed into a two-stage problem. The first stage is the strip combination problem aimed at determining the strip combination of each rolling turn and the number of rolling turns with the objective of minimizing the number of virtual strips, and the second is the strip allocation and sequencing problem aimed at optimizing the allocation and rolling sequence of the strips in each rolling turn. We first model this two-stage problem considering a set of production constraints and then design an optimal approach to solve the strip combination problem. Subsequently, we design an evolutionary algorithm (i.e., artificial bee colony algorithm) with a novel search strategy for employed bees, a dynamic strategy for onlooker bees, a variable neighborhood search strategy for a scout bee, and an enhanced strategy to solve the problem in the second stage. Computational experiments demonstrate the effectiveness of the proposed algorithms.Note to Practitioners—The hot rolling batch scheduling process is crucial in linking the casting and rolling processes of iron and steel productions. In the rolling batch scheduling problem of CSP, there is no buffer between the casting and rolling processes, and virtual strips must be added to satisfy production constraints. Most rolling batch scheduling methods do not consider the addition of virtual strips. In this paper, we mathematically characterize the hot rolling batch scheduling problem in CSP with flexible production constraints. We then show how the optimal approach and artificial bee colony algorithm are designed. Finally, the effectiveness of the proposed algorithms is demonstrated by comparisons with other well-known metaheuristic algorithms. This paper can be extended to other hot rolling batch scheduling problems with buffers and hybrid flowshop scheduling problems.

Operational indices optimization is crucial for the global optimization in beneficiation processes. This paper presents a multitasking multiobjective evolutionary method to solve operational indices optimization, which involves a formulated multiobjective multifactorial operational indices optimization (MO-MFO) problem and the proposed multiobjective MFO algorithm for solving the established MO-MFO problem. The MO-MFO problem includes multiple level of accurate models of operational indices optimization, which are generated on the basis of a data set collected from production. Among the formulated models, the most accurate one is considered to be the original functions of the solved problem, while the remained models are the helper tasks to accelerate the optimization of the most accurate model. For the MFO algorithm, the assistant models are alternatively in multitasking environment with the accurate model to transfer their knowledge to the accurate model during optimization in order to enhance the convergence of the accurate model. Meanwhile, the recently proposed two-stage assortative mating strategy for a multiobjective MFO algorithm is applied to transfer knowledge among multitasking tasks. The proposed multitasking framework for operational indices optimization has conducted on 10 different production conditions of beneficiation. Simulation results demonstrate its effectiveness in addressing the operational indices optimization of beneficiation problem. Note to Practitioners-Operational indices optimization is a typical approach to achieve global production optimization by efficiently coordinating all the indices to improve the production indices. In this paper, a multiobjective multitasking framework is developed to address the operational indices optimization, which includes a multitasking multiobjective operational indices optimization problem formulation and a multitasking multiobjective evolutionary optimization to solve the above-formulated optimization problem. The proposed approach can achieve a solution set for the decision-making. The simulation results on a real beneficiation process in China with 10 operational conditions show that the proposed approach is able to obtain a superior solution set, which is associated with a higher grade and yield of the product.

Gaussian processes (GPs) are the most popular model used in surrogate-assisted evolutionary optimization of computationally expensive problems, mainly because GPs are able to measure the uncertainty of the estimated fitness values, based on which certain infill sampling criteria can be used to guide the search and update the surrogate model. However, the computation time for constructing GPs may become excessively long when the number of training samples increases, which makes it inappropriate to use them as surrogates in evolutionary optimization. To address this issue, this paper proposes to use ensembles as surrogates and infill criteria for model management in evolutionary optimization. A heterogeneous ensemble consisting of a least square support vector machine and two radial basis function networks is constructed to enhance the reliability of ensembles for uncertainty estimation. In addition to the original decision variables, a selected subset of the decision variables and a set of transformed variables are used as inputs of the heterogeneous ensemble to further promote the diversity of the ensemble. The proposed heterogeneous ensemble is compared with a GP and a homogeneous ensemble for infill sampling criteria in evolutionary multiobjective optimization. Experimental results demonstrate that the heterogeneous ensemble is competitive in performance compared with GPs and much more scalable in computational complexity to the increase in search dimension.

Conventional evolutionary algorithms (EAs) are not well suited for solving expensive optimization problems due to the fact that they often require a large number of fitness evaluations to obtain acceptable solutions. To alleviate the difficulty, this paper presents a multitasking evolutionary optimization framework for solving computationally expensive problems. In the framework, knowledge is transferred from a number of computationally cheap optimization problems to help the solution of the expensive problem on the basis of the recently proposed multifactorial EA (MFEA), leading to a faster convergence of the expensive problem. However, existing MFEAs do not work well in solving multitasking problems whose optimums do not lie in the same location or when the dimensions of the decision space are not the same. To address the above issues, the existing MFEA is generalized by proposing two strategies, one for decision variable translation and the other for decision variable shuffling, to facilitate knowledge transfer between optimization problems having different locations of the optimums and different numbers of decision variables. To assess the effectiveness of the generalized MFEA (G-MFEA), empirical studies have been conducted on eight multitasking instances and eight test problems for expensive optimization. The experimental results demonstrate that the proposed G-MFEA works more efficiently for multitasking optimization and successfully accelerates the convergence of expensive optimization problems compared to single-task optimization.

Wireless networks attract increasing interests from a variety of industry communities. However, the wide applications of wireless industrial networks are still challenged by unreliable services due to severe multipath fading effects. Such effects are not only caused by massive metal surfaces but also moving operators and logistical vehicles, which will lead to temporal fading effects. A three-layer impulse response framework is proposed to characterize such effects, in which both the specular and scattered components vary with the spacial movement of nearby objects. In this context, a received signal strength indicator will be a noisy estimation only on the specular power and fail to describe the link quality accurately without the aid of scattered power. Consequently, an augmented Kalman-filter-based link quality estimator has been designed to track both the specular and scattered power in the distribution parameter space with constant noise covariance matrices. Experiments from industrial sites show significantly increased accuracy.

Surrogate models have shown to be effective in assisting metaheuristic algorithms for solving computationally expensive complex optimization problems. The effectiveness of existing surrogate-assisted metaheuristic algorithms, however, has only been verified on low-dimensional optimization problems. In this paper, a surrogate-assisted cooperative swarm optimization algorithm is proposed, in which a surrogate-assisted particle swarm optimization (PSO) algorithm and a surrogate-assisted social learning-based PSO (SL-PSO) algorithm cooperatively search for the global optimum. The cooperation between the PSO and the SL-PSO consists of two aspects. First, they share promising solutions evaluated by the real fitness function. Second, the SL-PSO focuses on exploration while the PSO concentrates on local search. Empirical studies on six 50-D and six 100-D benchmark problems demonstrate that the proposed algorithm is able to find high-quality solutions for high-dimensional problems on a limited computational budget.

Using evolutionary algorithms (EAs) to handle dynamic multiobjective optimization problems (DMOPs) is a challenging topic. In this paper, a prediction strategy based on reference points is proposed to improve the performance of EAs in solving DMOPs. The reference point based strategy is to partition the population into several subpopulations according to the reference points. When a change is detected, a sequence of the subpopulation centers in the previous environments belonging to the same reference point are used to estimate the center of in new environment. Based on the predicted centers, the EA will generate an initial population for the new environment using a combination of a Uniform distribution for enhancing population diversity and a Gaussian distributions for accelerating convergence. Experiments on ten test instances have been carried out to evaluate the performance of proposed strategy and the results show that reference point based prediction strategy exhibits superior performance in dealing with DMOPs with nonlinear correlations between decision variables and severe environmental changes.

This paper provides a method for automatically selecting optimal operational indices for unit processes in an industrial plant using measured data and without knowing dynamical models of the unit process. A dynamic multiobjective optimization problem is defined to find operational indices that lead to plant-wide production indices close to their target values. A case-based reasoning (CBR) technique is also employed, which uses the stored experience of a human expert to determine appropriate operational indices for given target production indices. The solutions of the optimization problem and CBR technique are combined to form baseline operational indices. The dynamic models of the production indices, however, are time varying and affected by disturbances and online corrections of these baseline operational indices are required. To this end, reinforcement learning (RL) is used to provide a data-driven optimization technique to compensate for disturbances and model approximation errors and variations. The data-driven RL approach is used in two different time scales. The samples of the predicted production indices are used at a fast sampling rate, i.e., at each sample time, and the samples of actual production indices are used at a slower sampling rate, i.e., after each operational run, to correct the baseline operational indices. The effectiveness of this automated decision procedure has been demonstrated by successful implementation of the proposed approach on a large mineral processing plant in Gansu Province, China.

In this paper, a novel pseudo gradient based DE approach is proposed, which takes advantage of both the differential evolutionary (DE) and the gradient-based algorithm. The gradient information, which is called pseudo gradient, is generated through randomly selected two vectors and their fitness function values. This work is to investigate the effect of proposed pseudo gradient on differential evolutionary algorithm. The simulation results show that DE with pseudo gradient can obtain better performance overall in comparison with classical DE variants. The pseudo gradient based DE with adaptive parameter section is compared with the existing adaptive DE algorithms. Also, the control parameter, step size are investigated to understand the mechanism of pseudo gradient in detail.

The whole production line of hematite ore processing is composed of raw ore processing, shaft furnace roasting, grindings, and magnetic separation production phases. Their automation systems consist of the process control part and the operational optimization system. The target of the optimal operational control is to optimize the concerned operational indices, namely, the intermediate product quality, efficiency, and consumptions. The dynamics between the operational indices and the global production indices (i.e., the total concentration grade, metal recovery rate, production rate, beneficiation ratio, and costs) with month, day, and hour time scales changes in line with the variations of production conditions, composition of raw ore together with capability of equipment. These indices are difficult to measure online and as a result it is difficult to model accurately. Moreover, there are characteristics in terms of both interconnections and conflictions among these indices. This leads to isolated operation of individual automation systems for these processes and the optimization of global production indices for whole production line cannot be realized. This paper presents a novel problem description for the integrated optimization of the automation systems of mineral processing. For this purpose, the analysis is made on the difficulty of using the existing optimization methods-based decision making methods to obtain the integrated optimization of the automation systems. The integrated optimization strategy for the automation systems of mineral processes is proposed using our previously established target value optimization of global production indices , two time scales decomposition approach and target value optimization of operational indices. The proposed strategy aims at realizing the optimization of global production indices. Using real data from a mineral processing plant on hematite beneficiation process, relevant simulations, and real industrial experiments have been carried out. The obtained experimental results show the efficiency and effectiveness of the proposed strategy.

For large electricity users, such as smelting plants, their electric loads cannot exceed a concerted limit in production. Traditional single-furnace optimization methods aim to satisfy the electric demand of a furnace to improve its production, and hence cannot consider the maximum demand constraint in a smelting plant. Maximum demand (MD) control is often utilized to keep the total electric demand within the limit via shedding the electric loads of some furnaces once the demand approaches the limit. However, the control method will enlarge the fluctuation of electric loads, which does harm to the production and causes a decline in energy-efficiency. In this paper, we propose a multifurnace optimization strategy to improve the production targets of a whole plant instead of a single furnace. In the strategy, an offline multiobjective load scheduling is first performed to assign electric loads for furnaces in each sampling period, taking into account of the MD constraint and production constraints. A multiobjective particle swarm optimization algorithm, combined with population initialization and constraint-handing strategies, is proposed to search for the Pareto optimal set of the scheduling problem, from which decision-makers can select one solution as the load scheduling program. A double closed-loop control mechanism is used to change the scheduled load into detailed load setpoints of furnaces and keep the actual loads up with the load setpoints. In the outer loop, the detailed load setpoints of furnaces are dynamically adjusted based on the deviation of actual loads from the scheduled loads. Thereafter, the desired setpoints are sent to the automatic control mechanism of each furnace, which is in the inner loop and responsible to keep the actual load up with the setpoint via a proportional-integral-derivative (PID) controller. The case study on a typical magnesia-smelting plant shows that the proposed multifurnace optimization strategy can achieve an increase of about 12.29% in the production output, an improvement of about 0.46% of the magnesia in the product, and a slight reduction of 2.35% in electricity cost over the results of MD control. Note to Practitioners - For large electricity users, such as smelting plants, they are subjected to the maximum electric demand constraint. The maximum demand control device is widely adopted to solve the problem, but it will cause a decline in production output and energy-efficiency. This paper was motivated by improving the multiple production targets (i.e, the total production output, the product quality, and the total electricity cost) of a plant via load scheduling and control. In contrast to the maximum demand control, the load scheduling and control approach is a beforehand strategy that can optimize the operation. A case study on a magnesia-smelting plant shows that the proposed approach performs better than the maximum demand control technique.

In this paper, a novel knowledge-based global operation approach is proposed to minimize the effect on the production performance caused by unexpected variations in the operation of a mineral processing plant subjected to uncertainties. For this purpose, a feedback compensation and adaptation signal discovered from process operational data is employed to construct a closed-loop dynamic operation strategy. It uses the signal to regulate the outputs of the existing open-loop and steady-state based system so as to compensate the uncertainty in the steady-state operation at the plant-wide level. The utilization mechanism of operational data through constructing increment association rules is firstly described. Then, a rough set based rule extraction approach is developed to generate the compensation rules. This includes two steps, namely the determination of the variables to be compensated based on the significance of attributes in the rough set theory and the extraction of the compensation rules from process data. Based upon the operational data of the mineral processing plant, relevant rules are obtained. Both simulation and industrial experiments are carried out for the proposed global operation, where the effectiveness of the proposed approach has been clearly justified.

This paper presents a novel offline modeling for product quality prediction of mineral processing which consists of a number of unit processes in series. The prediction of the product quality of the whole mineral process (i.e., the mixed concentrate grade) plays an important role and the establishment of its predictive model is a key issue for the plantwide optimization. For this purpose, a hybrid modeling approach of the mixed concentrate grade prediction is proposed, which consists of a linear model and a nonlinear model. The least-squares support vector machine is adopted to establish the nonlinear model. The inputs of the predictive model are the performance indices of each unit process, while the output is the mixed concentrate grade. In this paper, the model parameter selection is transformed into the shape control of the probability density function (PDF) of the modeling error. In this context, both the PDF-control-based and minimum-entropy-based model parameter selection approaches are proposed. Indeed, this is the first time that the PDF shape control idea is used to deal with system modeling, where the key idea is to turn model parameters so that either the modeling error PDF is controlled to follow a target PDF or the modeling error entropy is minimized. The experimental results using the real plant data and the comparison of the two approaches are discussed. The results show the effectiveness of the proposed approaches.