With the increasing complexity of industrial production systems, accurate fault diagnosis is essential to ensure safe and efficient system operation. However, due to changes in production demands, dynamic process adjustments, and complex external environmental disturbances, multiple operating conditions frequently arise during production. The multi-condition characteristics pose significant challenges to traditional fault diagnosis methods. In this context, multi-condition fault diagnosis has gradually become a key area of research, attracting extensive attention from both academia and industry. This paper aims to provide a systematic and comprehensive review of existing research in the field. Firstly, the mathematical definition of the problem is presented, followed by an overview of the current research status. Subsequently, the existing literature is reviewed and categorized from the perspectives of single-model and multi-model approaches. In addition, typical real-world application scenarios are then summarized and analyzed. Finally, the key challenges and prospects in the field are thoroughly discussed.

Real-time safety assessment is a critical process for identifying and analyzing potential safety hazards in industrial applications. Active learning has been widely recognized as an effective technique for addressing such issues by utilizing small labeled samples to achieve high evaluation performance. However, in real-world scenarios, the performance of query strategies can be significantly impacted by the data environment. Therefore, it is crucial to co-design a strategy using different criteria to ensure reliable and stable safety assessment results for system operation. In this paper, we propose a contrastive preference-guided active learning approach to tackle chunk-level real-time safety assessment tasks in non-stationary environments. Firstly, we construct negative ranking lists and random lists. Then, we introduce Jeffrey divergence to measure pairwise ranking correlation. By leveraging contrastive preference relationships, we can effectively obtain the value of samples in data chunks with preference aggregation procedures. To verify the effectiveness of the proposed method, we conduct numerous experiments using realistic data from the JiaoLong deep-sea manned submersible. The results demonstrate that our approach outperforms most existing advanced methods in terms of performance stability and accuracy.Note to Practitioners—In situations where a significant number of samples must be processed concurrently to promptly identify and mitigate potential hazards, it is vital to consider the responsibilities of real-time safety assessment (RTSA). The proposed approach offers a resolution to the difficulty of acquiring annotations for all samples due to the continual stream of data. Multiple query criteria can be seamlessly integrated, which enhances the stability and superiority of learning performance simultaneously. The approach’s scalability, flexibility, and effectiveness render it a valuable instrument for guaranteeing a prompt response to emergent procedural threats and safety. It can accommodate any number and type of advanced single query criteria, making it easy to adapt to various scenarios. The proposed CPRC approach signifies a significant breakthrough in the field of real-time safety assessment, with the potential to enhance safety outcomes in industrial settings. Multiple experiments involving the realistic JiaoLong deep-sea manned submersible were conducted, and the outcomes demonstrate the benefits of this approach for practical RTSA applications.

Effective fault diagnosis is crucial for maintaining the reliability and safety of industrial systems. Incremental learning, which enables models to continuously update and adapt to new data or emerging fault classes without complete retraining, has recently gained attention as a promising solution for addressing nonstationary data streams in fault diagnosis applications. Nevertheless, most existing review articles on fault diagnosis adopt a broad perspective, primarily discussing general techniques such as deep learning and transfer learning, without providing a dedicated focus on incremental learning strategies. To the best of our knowledge, it is the first review focusing specifically on incremental learning-enabled fault diagnosis methods. In this work, state-of-the-art incremental learning-enabled fault diagnosis are systematically reviewed. These methods are categorized into distinct groups based on their incremental learning strategies and application contexts. In addition, major challenges associated with applying incremental learning to fault diagnosis, including concept drift and catastrophic forgetting, are discussed, along with emerging solutions proposed to address these issues. A novel taxonomy and perspective on incremental learning-enabled fault diagnosis approaches is presented, providing a timely and comprehensive reference for researchers and practitioners in this evolving field.

Real-time safety assessment (RTSA) of dynamic systems holds substantial implications across diverse fields, including industrial and electronic applications. However, the complexity and rapid flow nature of data streams, coupled with the expensive label cost and pose significant challenges. To address these issues, a novel confusion-based learning framework, termed confusion-and-detection method plus (CADM+), is proposed in this article. When drift occurs, the model is updated with uncertain samples, which may cause confusion between existing and new concepts, resulting in performance differences. The cosine similarity is used to measure the degree of such conceptual confusion in the model. Furthermore, the change of standard deviation within a fixed-size cosine similarity window is introduced as an indicator for drift detection. Theoretical demonstrations show the asymptotic increase of cosine similarity. In addition, the approximate independence of the change in standard deviation with the number of trained samples is indicated. Finally, the extreme value theory (EVT) is applied to determine the threshold of judging drifts. Several experiments are conducted to verify its effectiveness. Experimental results prove that the proposed framework is more suitable for RTSA tasks compared with state-of-the-art algorithms. The source code is available at https://github.com/THUFDD/CADM-plus.

In response to the increasing complexity of tasks in artificial intelligence, broad learning systems (BLSs) have emerged as essential tools, especially given the limitations of deep neural networks, such as their extensive training and computational demands. This study addresses the computational inefficiencies and numerical instabilities inherent in traditional BLS when handling complex tasks in dynamic environments. To mitigate these challenges, we propose an enhanced version of BLS incorporating QR factorization (QRF), referred to as QRBLS, which is known for improving numerical stability. This framework replaces the traditional method of computing output weights, which typically relies on the Moore-Penrose pseudoinverse. The primary contribution of this article is the integration of QRF into the BLS architecture, thereby improving stability when processing large-scale datasets. QRBLS also features a dynamic updating mechanism that adjusts model parameters efficiently with new data, enabling continuous learning without the need for full-model re-evaluation. In addition, a time-dependent structure (TDS) enhances the model's responsiveness to temporal data changes, increasing its utility in dynamic environments. Validation through numerical experiments demonstrated that QRBLS outperformed traditional BLS, exhibiting superior stability and adaptability in handling data anomalies and rapid updates. The integration of QRF and TDS significantly improves the adaptability and computational efficiency of BLS, providing a robust solution for large scale and dynamic AI applications. QRBLS effectively addresses challenges related to numerical instability and continuous learning, offering practical improvements in real-world settings.

Diagnosis of compound faults is meaningful and challenging in actual industrial applications. Generally, it is impractical to obtain a sufficient amount of labeled data for the compound faults, which limits the performance of existing methods. The characteristics of compound faults should be fully considered. Hence, reasonably introducing prior knowledge of fault information is of great significance to improve the practicability of diagnostic methods. In this paper, several unsupervised representation extraction techniques are firstly exploited to extract the fault information in the latent space, which is beneficial to alleviate the negative effects of noise disturbance in real-life scenarios. A discrimination-guided active learning method based on marginal representations, termed DGMR, is then proposed. In this case, samples with compound fault information are more likely to be iteratively queried for expert annotation. Several experiments are conducted using realistic experimental platform data. The experimental results show that the proposed DGMR can achieve high diagnostic accuracy for industrial compound fault diagnosis with a small number of annotated samples and shallow classifiers.Note to Practitioners—Compound faults widely exist in complex industrial equipment due to device coupling, which tends to be more diverse and generally exhibits more complex characteristics. In addition, even if the device passes the consistency test, the distribution of data collected by the test device before and after long-term use may have certain deviations due to performance degradation and other factors. Moreover, expert annotation is necessary but cost-sensitive. With the proposed scheme in this paper, vibration signals can be used to construct a fault database through several feature extraction techniques. Compound and unknown fault samples are more likely to be selected. Engineers can provide labels for these small number of fault samples based on methods such as frequency domain analysis, which can reduce the cost of annotations. High diagnostic accuracy can be obtained only using shallow classifiers in this case. Several experiments of a real rotating machinery fault diagnosis test rig are carried out. Experimental results demonstrate that the proposed method outperforms some advanced methods.

Chunk-level real-time safety assessment of dynamic systems is a critical component of industrial processes, which is essential to prevent hazards and reduce the risk of injury or damage to equipment and facilities, especially in nonstationary environments. In this context, multiple real and complex concept drifts are inevitable in industrial settings, making it crucial to understand their detection and adaptation processes. The incremental learning scheme should also be well considered. However, existing methods have certain limitations in dealing with such issues. In this article, a dynamic model interpretation-guided online active learning scheme, termed a dynamic model interpretation-guided learning scheme (DMI-LS), is proposed. Specifically, the model update strategy with chunk data is designed based on the implementation of the broad learning system. A novel query strategy is then investigated to consider the ranking preference difference, which relies on the interpretation generated by the explainable artificial intelligence method. Several experiments based on the JiaoLong deep-sea manned submersible data are conducted to verify the effects of the proposed DMI-LS. The results show that it outperforms the other advanced existing approaches with different settings in most scenarios.

Operational changes in industrial production can alter system operating modes, which complicates real-time fault diagnosis by affecting sensor data and fault characteristics. In addition, fault diagnosis tasks encounter the challenge of fault feature drift, which causes a decline in the performance of previously trained models on new data. This article presents a novel approach for real-time multimode fault diagnosis called the evidential ensemble preference-guided approach to tackle these issues. During the offline stage, we extract ensemble preferences of fault information across different operating modes based on the structure of the broad learning system. Subsequently, a parameter iterative update rule is developed that utilizes an evidential reasoning technique to emphasize the preferences during the online stage. The effectiveness of our approach is evaluated by constructing a real-time multimode fault diagnosis dataset using the Tennessee Eastman process and conducting multiple experiments. The results demonstrate that our proposed approach effectively identifies operating modes and diagnoses faults simultaneously, surpassing existing advanced methods.

Real-time safety assessment of dynamic systems is of paramount importance in industrial processes since it provides continuous monitoring and evaluation to prevent potential harm to the environment and individuals. However, there are still several challenges to be resolved due to the requirements of time consumption and the non-stationary nature of real-world environments. In this paper, a novel online dynamic hybrid broad learning system, termed ODH-BLS, is proposed to more fully utilize the co-design advantages of active adaptation and passive adaptation. It makes effective use of limited annotations with the proposed sample value function. Simultaneously, anchor points can be dynamically adjusted to accommodate changes of the underlying distribution, thereby leveraging the value of unlabeled samples. An iterative update rule is also derived to ensure adaptation of the assessment model to real-time data at low computational costs. We also provide theoretical analyses to illustrate its practicality. Several experiments regarding the JiaoLong deep-sea manned submersible are carried out. The results demonstrate that the proposed ODH-BLS method achieves a performance improvement of approximately 8% over the baseline method on the benchmark dataset, showing its effectiveness in solving real-time safety assessment tasks for dynamic systems.

The design of real-time safety assessment (RTSA) approaches in nonstationary environments is meaningful to reduce the possibility of significant losses. However, several challenging problems are needed to be well considered. The performance of existing approaches will be negatively affected in the settings of imbalanced drifting streams. In this case, the model design with the incremental update should also be explored. Furthermore, the query strategy should also be well-designed. This article investigates a dynamic submodular-based learning strategy to address such issues. Specifically, an efficient incremental update procedure is designed with the structure of the broad learning system (BLS), which is beneficial to the detection of concept drift. Furthermore, a novel dynamic submodular-based annotation with an activation interval strategy is proposed to select valuable samples in imbalanced drifting streams. The lower bound of annotation value is also proven theoretically with a novel drift adaption mechanism. Numerous experiments are conducted with the realistic data of JiaoLong deep-sea manned submersible. The experimental results show that the proposed approach can achieve better assessment accuracy than typical existing approaches.

The degradation of a system's performance poses a significant challenge to the effective application of fault diagnosis methods for dynamic systems. Consequently, the underlying feature distribution changes over time during the actual process, resulting in a decline in the effectiveness of existing diagnosis methods. In this article, we present a real-time adaptive fault diagnosis scheme to address this issue. A latent variable-guided broad learning system (LVGBLS) is proposed to construct the fundamental diagnosis model, which effectively extracts dynamic features from the monitored data. An incremental update procedure is then designed based on pseudolabel learning to adapt to dynamic process changes while minimizing labeling costs. We also introduce the condition detection mechanism (CDM) to detect dynamic changes under the degradation process based on statistical information. To demonstrate the effectiveness of our proposed method, we conduct several comparison experiments and ablation experiments on electrical drive systems and XJTU-SY bearing datasets. The results show that our proposed scheme exhibits superior performance with low labeling costs in most scenarios with performance degradation.

Multisource data fusion analysis, particularly in decision-level fusion strategies, is emerging for application in real-life scenarios. The Dempster–Shafer evidence theory (DSET) is a prevalent approach that has significant importance in managing the fusion tasks. However, existing fusion approaches have limitations in dealing with redundant information and computational complexity associated with the fusion procedure. Though conflict management has been thoroughly studied, other limitations have not been well addressed. In this article, we propose a novel approach for evidential multisource data fusion based on game-theoretic analysis. The introduction of the Shapley function considers the interaction effect of focal elements, mitigating the negative influence of redundant evidence. Additionally, the computational complexity of the fusion procedure is reduced to the same level as the approximate Bayesian update model. We provide a numerical example with conflicting and redundant evidence to show that the proposed approach outperforms current advanced weighted average-based fusion methods. Moreover, a simulation experiment demonstrates the practicality and effectiveness of the proposed approach in identifying driver fatigue states based on electroencephalography (EEG) signals.

Real-time safety assessment of dynamic systems has recently received increasing attention. However, the performance of existing advanced approaches is often negatively affected by realistic requirements such as limited annotations and memory. In this case, how to design reasonable query strategies to select valuable instances and exploit the memory space efficiently is extremely meaningful. This paper proposes a novel memory-triggered submodularity-guided active broad learning approach, termed MTSGABL, to deal with such issues simultaneously. Specifically, the broad learning system is introduced as the basic assessment model to update incrementally. A memory-triggered learning mechanism is then proposed based on the drift detection procedure, which controls the update process to exploit the latest sequential information. Furthermore, a submodularity-guided query strategy is introduced to select a small number of valuable samples sequentially, which is beneficial to alleviate the negative effects of the imbalanced data stream. Numerous comparison and ablation experiments with the realistic JiaoLong deep-sea manned submersible data are conducted to validate its effectiveness. Results show that the proposed approach is superior to the existing typical approaches subject to these constraints.

Real-time safety assessment of the complex dynamic systems in nonstationary environments is of great significance for avoiding the potential hazards. In this case, the update procedure with high assessment accuracy and training speed is crucial and meaningful in the dynamic streaming setting. Generally, the performance of most online learning approaches will be negatively affected by limited annotated samples in such a setting. Moreover, the time cost of advanced conventional methods with retaining procedures is relatively high, constraining their practicality. In this article, a novel online active broad learning approach, termed OABL, is proposed. In detail, the effectiveness of the broad learning system in the framework of online active learning is first revealed and verified. A reasonable dynamic asymmetric query strategy is then designed with a limited annotation budget to actively annotate the relatively valuable samples, which is beneficial to mitigating the negative effects of class imbalance. In this context, the advantage of the human-in-the-loop characteristic is also effectively used to control the evolution direction of the learner during the incremental update, which makes it better able to adapt to complex and nonstationary environments. Several related experiments are conducted with the realistic data of JiaoLong deep-sea manned submersible. Results show the effectiveness and practicality of the proposal compared with the existing advanced approaches.

The consideration of social influence in solving group decision-making (GDM) problems has caused widespread concern in recent years. In most cases, the interaction of experts and the association efforts of alternatives are often not to be ignored in the actual decision-making environment, especially for those situations with incomplete information in the process of opinion changes. In this study, a novel measure-based GDM model is proposed to more comprehensively consider such a factor. Also, a positive cooperation-first principle is proposed as a guide for situations with incomplete information. One of the main advantages of the proposed model is that the estimation of missing fuzzy preference relations is completed according to the modeling results of peer interactions, which makes the preference evolution results more reliable. With the utilization of the fuzzy measure framework, the Shapley function and interaction indicator, originated from the cooperative game theory, are also introduced to extract the interaction features. After several iterations, the global opinion can be eventually obtained via the social influence network technique. Furthermore, it is also demonstrated that the evolution of individual opinions converges to the ultimate collective opinion. A case study of supplier selection with the proposed model is implemented to illustrate its practicality and effectiveness. Related comparisons and discussion regarding related methods are also mentioned.

Multicriteria decision analysis evaluates multiple conflicting criteria in decision making, but conflicting criteria are typical in evaluating options. As the existing ordering operations involved in multicriteria decision making cannot easily be implemented with intervals, we assume that scalar representative values with intervals can effectively avoid this issue. To deal with interval-valued criteria, we propose a generalized golden rule representative value approach, which involves the sigmoid function of backpropagation neural networks to tune parameters. Our approach considers the uncertainties and side effects of the interval variables to improve individual scalar representative values. Based on numerical examples, we address the effectiveness of the proposed approach, and we provide a specific application concerning multicriteria decision making with interval criteria satisfaction.

Multicriteria decision-making problems have been applied to many applications for its practicality. Nevertheless, when the evaluated satisfactions are more complex, such as interval-valued distributions, how to reasonably obtain the aggregation results of alternatives is still an open issue. In this paper, an interval-valued exceedance method is proposed to solve such a question based on the Golden Rule representative value and probabilistic exceedance method. Due to good performance of expressing uncertain information, the Golden Rule representative value method is used to order interval-valued satisfactions after an effective normalization process. In addition, a quantifier-based ordered weighted averaging operator is also introduced to consider the preferences of decision makers. A realistic application of supplier selection is shown to illustrate the practicality of the proposed method.