Due to the extensive interconnectedness of components in modern engineering systems, fault diagnosis plays a pivotal role in ensuring system reliability, operational continuity, and proactive maintenance. Prediction accuracy becomes crucial in fault diagnosis, serving as a key metric to identify specific faulty conditions and enabling timely and precise maintenance operations. In recent years, ensemble learning has emerged as an effective branch of machine learning, demonstrating the capability to achieve accurate diagnosis and classification. This study presents a novel hierarchical ensemble learning approach based on a two consecutive learning phases. The first stage employs a bagging technique that utilizes multiple integrated ensembles, each containing a different set of base classifiers. The second phase employs a stacking technique to improve accuracy and reliability of the fault diagnosis process. We validate the feasibility and effectiveness of the proposed approach for cable network systems impairments identification using real DOCSIS® Full Band Capture (FBC) downstream data, achieving state-of-the-art classification accuracy for fault categories derived from the FBC data sources. This represents the first method in the literature addressing the diagnosis of such systems using this type of data. Evaluation methodology introduces classification experiments performed incrementally, starting with binary classification and progressing sequentially to the nine-class impairment diagnosis. Experimental results highlight the potential for enhanced fault detection and diagnosis capabilities, contributing to accurate fault diagnosis in cable telecommunication systems.

In industrial scenarios, mechanical faults are episodic and uncertain. Thus the monitoring data collected is usually extremely imbalanced, resulting in intelligent diagnostic models that suffer from majority-class dominance, minority-class overfitting, and poor generalization performance. Therefore, a knowledge distillation-guided cost-sensitive ensemble learning framework is proposed. It effectively combines ensemble learning and cost-sensitive learning to fully extract the multiscale features, effectively leverage the critical multi-depth features, and emphasize classifying the most confusing classes. Specifically, multiple-scale feature extraction and multi-order fusion are first employed to fully utilize the fault information. Afterward, the complementary diagnostic knowledge at different depths of the network is embedded into a novel ensemble learning process for better integration decisions. Then an improved knowledge distillation method achieves the mutual transfer and sublimation of excellent diagnostic knowledge while focusing on the most confusing fault classes to achieve the effective representation of various types of faults. Finally, a cost-sensitive strategy is applied to further increase attention to minority classes. The experimental results for various complex data imbalance scenarios, including extreme imbalance, step imbalance, continuous imbalance, interclass imbalance, and intra-class imbalance, all indicate that the proposed method can achieve state-of-the-art performance and provide a promising solution for the practical industrial application of intelligent diagnostic methods.

As the IEEE Reliability Society proudly commemorates its 75th anniversary, we find ourselves at a pivotal juncture—a moment for reflection and a forward-looking celebration, particularly for our flagship publication,IEEE Transactions on Reliability. Over the decades, this esteemed journal has been at the forefront of advancing reliability engineering, shaping the way we perceive and ensure the dependability of complex systems in our rapidly evolving and technology-centric world.

The cyber-physical power system (CPPS) stands as one of a nation's most critical infrastructures, as an unwavering and secure power supply forms the cornerstone of national and societal development. Recently, the concept of resilience has become a trending topic in preventing and mitigating the risks caused by large-scale CPPS blackouts. This article introduces the concept of resilience in complex engineering systems, detailing quantitative assessment metrics used to evaluate CPPS resilience. These metrics underpin the capacity of CPPSs to recover from disruptions. The article reviews current research focused on enhancing CPPS resilience, covering three main aspects: 1) optimizing component recovery sequences, 2) identifying critical nodes, and 3) optimizing physical-cyber coupling patterns. Recent advances in modeling methodologies for CPPS resilience against cascading failures are also discussed. Finally, the article delves into the challenges and future direction for CPPSs resilience enhancement and modeling approaches.

Frequency security is critical for the power systems’ stability and reliability. The integration of renewable energy resources brings new challenges to the frequency security of power systems. To provide a more robust control method for controlling the frequency and tie-line power flow of an interconnected power system integrated with wind farms (IPSWF), a load frequency control (LFC) controller of generalized predictive control (GPC) based on the Takagi-Sugeno (T–S) fuzzy model (fuzzy-GPC) is proposed in this paper. First, the T–S fuzzy model of the IPSWF is constructed. Then, the GPC is used to design the controller. Lastly, to validate the proposed scheme, an LFC strategy is established for a two-area interconnected power system integrated with a wind farm in power systems computer aided design (PSCAD). The control performance of proportional-integral, GPC, and fuzzy-GPC controller are compared under two faulty operating conditions. The simulation results show that the performance indicators and the dynamic behavior when the fuzzy-GPC controller is used are better than the performance when the PI controller and GPC controller are used. The proposed fuzzy-GPC used in the LFC of the IPSWF can regulate the frequency deviation and tie-line power deviation adaptively and achieve minimum frequency deviation and tie-line power deviation in a multi-area IPSWF.

There are multiple warehouses in a multiechelon inventory system, and the size of the state space increases exponentially with the number of warehouses. Therefore, the curse of dimensionality becomes unavoidable when performing steady-state analysis. Most existing studies calculate the inventory cost or supply chain reliability based on specific assumptions. For example, it often assumes that the lead time is either zero or an integral multiple of the review period, and that each warehouse adopts a base-stock policy. This article considers a more practical and prevalent situation where the lead time is less than a review period, and a more general (s, S) strategy is adopted. The curse of dimensionality during steady-state analysis is alleviated by decomposing transition probabilities. Then, the cost and supply chain reliability are derived from steady-state distributions. Finally, a case study involving spare part inventory of wind turbines is considered. Nondominated inventory strategies are obtained using the particle swarm optimization method to strike a balance between costs for the wind turbine manufacturer and wind farm owners.

The papers in this special section focus on increasing interests in the development and implementation of advanced artificial intelligence (AI) and machine learning (ML) methods for tackling the reliability and system health prognostics challenges in various industrial applications.

The dynamic performance of a grid-connected inverter in a distributed generation system brings new challenges by affecting the power quality and dynamic stability. The traditional linear control method for the inverter requires complex processes including decoupling and control parameter tuning and relies on a pulsewidth modulation module. The nonlinear control method, e.g., the model predictive control (MPC), can address some of the aforementioned challenges but is incapable of dealing with system parameter changes. A data-driven method using the dynamic Bayesian network-based model predictive control (DBN-MPC) is proposed, which takes advantage of the predictive ability of the DBNs to implement the MPC strategy. A predictive model is built, and the prediction signals are generated based on the DBNs. Then, by constructing the cost function and optimization criteria, the controller generates the optimal switching state combination. Using the proposed DBN-MPC method, the predictive model implements parameter learning online, providing more accurate prediction signals for further optimization and the feedback correction. In addition, the control law of the proposed controller can update over time, which enables the grid-connected inverter system to achieve the optimal control. The case studies using the grid-connected inverter system and IEEE 39-bus benchmark power system integrated with the battery energy storage system demonstrate and verify the superiority of the proposed DBN-MPC method.

Failure mode and effects analysis (FMEA) has been widely used in product design process as a reliability analysis technique. Design FMEA (DFMEA), which is used in the product development phase to identify and mitigate product risk, is one of the three application scenarios of FMEA. During the DFMEA process, verification and validation (V&V) activities are proposed to mitigate the risk of the identified potential failure modes. The V&V activities can be further planned and implemented to improve the product reliability under development. However, the DFMEA report usually contains rich text descriptions of potential failure modes and causes, and it is difficult and nonintuitive to fully understand these information for design improvements. In addition, it is also very challenging to optimize the planning of V&V activities by selecting a set of V&V activities to achieve expected reliability improvement effectiveness with the minimum resource consumption requirement. To address these two challenges, this article first proposes a method of applying text mining to the DFMEA report to obtain two types of hidden reliability information, including the classifications of failure modes/causes and the correlation between keywords. Then, a mathematical model is proposed to optimize the product V&V planning by selecting an optimal set of V&V activities. The application of the above proposed methods is illustrated through the product development of a diesel engine power generation system.

This article investigates optimal recovery strategy of components for maximizing the resilience of the cyber–physical power system (CPPS), where a component represents a unique node or branch, such as generating stations and communication transmission lines. The proposed optimization model is built as a multimode resource-constrained project scheduling problem to incorporate system resilience, cascading failures of the CPPS, the diversity of recovery resources, execution modes of recovery activities, precedence of damaged components, as well as the availability, cost, and time of recovery resources. The failure propagation mechanisms are characterized by a cascading failure model, which is further embedded in the optimization model to quantify the system real-time performance during the recovery process, and determine whether repaired components can be reconnected to the system. The system resilience is quantified using a proposed time-dependent annual composite resilience metric based on a compound Poisson process. The proposed optimization model is solved using a modified simulated annealing algorithm. The system that couples the IEEE 30-bus model and a small-world communication network is used as a testbed to demonstrate the feasibility and effectiveness of the proposed modeling approach. Comparisons with existing optimization models in the literature show the superiority of the proposed model.

Due to the advancements in sensing technologies and computational capabilities, system health assessment and prognostics have been extensively investigated in the literature. Industry has adopted and implemented many advanced system prognostic applications. This article reviews recent research advances and applications in prognostics modeling methods for engineering systems. The reviewed papers are classified into three major areas based on whether the physics of failure knowledge is incorporated for prognostics, i.e., the data-driven, physics-based, and hybrid prognostic methods. The technical merits and limitations of each prognostic method are discussed. This review also summarizes research and technological challenges in engineering system prognostics, and points out future research directions.

This paper proposes an algorithm for evaluating the performance of a closed-loop spare machines supply system when the time between failures as well as the repair times is generally distributed. The closed-loop system with finite number of machines and the generic time distribution assumption well represents real-life spare machines supply systems but at the cost of higher complexity. The increased complexity prohibits the use of existing models, which evaluate the performance of closed-loop systems assuming exponentially distributed failure and repair times. The Extended Bottleneck (EBOTT) algorithm in the literature estimates the performance measures of a closed-loop system with generally distributed repair service times. However, as our analysis indicates, the EBOTT algorithm may fall into an infinite loop even in assessing the performance of a simple spare machines supply system. The Modified EBOTT algorithm presented in this paper overcomes the shortcomings of the EBOTT algorithm and exhibits superior performance. Finally, through tradeoff analysis between spare machines inventory level and repair shop capacities, insights into the performance of such systems are presented.

The seven papers included in this special section focus on reliability, resilience, and prognostics modeling of complex engineering systems. This topic has become very challenging due to factors such as complex physical interdependence and functional interactions among mechanical, electrical, software, and control subsystems, as well as the interactions with external environments. The difficulties of modeling the reliability and resilience as well as system performance prognostics of complex engineering system have been observed from many perspectives, e.g., 1) the schedule delays and cost overruns of major government and industry commercial programs, and 2) the challenges when planning and managing recovery activities of critical infrastructure damaged due to extreme events. When multiple sources of information and disparate data are available, how to monitor, model, and predict the system resilience and reliability over time of complex engineering systems becomes essential and critical for safe, reliable, resilient, and economic operations of most cyber-physical systems.

Failure mode and effect analysis (FMEA) is a well-known engineering technique to recognize and reduce possible failures for quality and reliability improvement in products and services. It is a group-oriented method usually conducted by a multidisciplinary and cross-functional expert panel. In this paper, we explore two key issues inherent to the FMEA practice: the representation of diversified risk assessments of FMEA team members and the determination of priority ranking of failure modes. Specifically, a framework integrating cloud model, a new cognitive model for coping with fuzziness and randomness, and preference ranking organization method for enrichment evaluation (PROMETHEE) method, a powerful and flexible outranking decision making method, is developed for managing the group behaviors in FMEA. Moreover, FMEA team members' weights are objectively derived taking advantage of the risk assessment information. Finally, we illustrate the new risk priority model with a healthcare risk analysis case, and further validate its effectiveness via sensitivity and comparison discussions.

This paper proposes a new multiobjective multiple stage reliability growth planning (MO-MS-RGP) model. The model is based on multiobjective consideration of developing a new product, including the cost, time, and product reliability. The number of test units, test time, and the percentage of introduced new technologies are considered as decision variables in the model. Varying reliability growth rates are considered for each subsystem in each stage. Product new technologies or contents can be completely introduced in one stage or partially introduced to the product over multiple stages. New product development time limit and budget are considered as constraints in the MO-MS-RGP model. An integrated approach is developed to formulate and solve the proposed MO-MS-RGP problem. The approach starts with a multiobjective evolutionary algorithm, called multipleobjective particle swarm optimization to find a set of Pareto optimal solutions. Then, clustering methods are applied to cluster the solutions obtained by the evolutionary algorithm. Finally, the clustered solutions are ranked using a multiple criteria decision making method. A numerical example illustrates the application of the proposed MO-MS-RGP model for the reliability growth planning optimization of a next generation engine development.

This paper proposes a multi-objective multi-stage reliability growth planning method in the early product-development stage. Multi-stage reliability growth planning is common in practice, and it aligns well with multiple developmental stages of a new product such as concept design, detail design, prototype design, and final production version design. The multi-objective formulation reflects the needs of product development's multiple objectives, such as program cost, schedule, and reliability. Pareto optimal solutions of the multi-objective multi-stage formulation for reliability growth planning are searched using a modified nondominated sorting genetic algorithm (NSGA-II). To reduce the large size of Pareto optimal solutions to a workable size of efficient solutions for plan implementation, both constant return-to-scale and variable return-to-scale data envelopment analysis (DEA) methods are used for determining the efficient solutions. Based on tradeoff and sensitivity analysis, insights and guidelines are presented for choosing appropriate reliability growth plans in terms of optimal allocation of testing time and testing units, and the timing for new technology introduction. The growth rate in each product-development stage and its impact on the development cost, schedule, and reliability are also discussed. An illustrative example is given to demonstrate the approach for planning the reliability growth for a next-generation engine development.

This paper proposes a new perspective and methodology to model the behavior of the system/component using the theory and methods for fuzzy sets. We use the indicator or performance or substitute variable which is well understood by the customer to fuzzify the states of the system. Thus, the success/failure events are treated as fuzzy sets. Fuzzy reliability is defined and compared to the traditional binary states and multi-state reliability modeling. The concept of fuzzy random variable is introduced to model the dynamic behavior of time to fuzzy failure. Customer-centric system performance measures are developed based on fuzzy reliability modeling. Numerical examples show how fuzzy reliability models and relevant measures can be applied to evaluate complex system and the advantages over the classic reliability analysis in decision making.