Proceedings of the Institution of Mechanical Engineers,Part O: Journal of Risk and Reliability

Given the current strict regulations on fuel emissions, hydrogen fuel cell vehicles have been considered as ideal non-polluting vehicles. As the kernel subsystem of hydrogen fuel cell vehicles, the fuel cell engine, which is a complex system, may fail because of multiple competing failure causes. In fuel cell engine reliability engineering, discriminating and measuring dependency among multiple failure causes is an urgent problem that should be addressed. This study proposes a statistical dependency analysis model under competing risks based on reliability field data of fuel cell engines. The multivariate lognormal distribution is employed to model joint failure distribution. Then, using simulated annealing algorithm to estimate the parameters of the conditional probability likelihood function. Moreover, p-value hypothesis test procedures are developed to determine the significance of the dependence degree among multiple competing failure causes. In this article, the failure mechanisms of fuel cell engine are comprehensively discussed to prove the feasibility and effectiveness of the proposed method. Results can provide technical support for studies on optimal maintenance strategies.

The terms ‘accuracy’ and ‘precision’ are commonly used in the scientific literature to describe the quality of, for example, measurements or methods, where traditional definitions of the terms often associate ‘accuracy’ with systematic errors and ‘precision’ with random errors. However, several different definitions exist, and the meaning of the terms may vary. In a risk analysis context, which we study in more detail in this article, it is common to discuss new methods and risk results using the terms ‘accuracy’ or ‘precision’, but, in many situations, a proper clarification of what these terms actually mean is missing. Many authors seem to mix or not differentiate between the terms, some assuming that they are synonymous and a contrasting of the terms is unnecessary. As the concept of risk may also be subject to several interpretations, this may cause confusion and lead to unintended conclusions. A simple example from the oil and gas industry is used to illustrate the situation. The article offers some clarification of the current use of the terms and gives specific suggestions for how to interpret them in risk analysis. An outcome of the article is recommendations on how to define the terms in a risk analysis context.

Reliability modeling of repairable systems deals mostly with two types of repair. Perfect repair brings a system to “as good as new” state. Minimal repair, on the contrary, returns a system to the state immediately prior to failure. In this article, we consider perfect and imperfect preventive maintenance actions for a system subjected to minor and major failures. Minor failures are minimally repaired, whereas a major failure terminates the operational function of the system and can be considered as an end-of-life event. The preventive maintenance strategies that we propose and analyze increase mission success probability and extend the expected lifetime of the system. The modeling is illustrated with numerical examples.

This article extends the finite Markov chain imbedding approach to evaluate the system reliability of a dynamic k-out-of-n:F system operating under two cyclic alternating conditions, and this article presents a method to obtain the optimal replacement interval of the system according to age-based replacement policy. In terms of the “dynamic k-out-of-n:F” model, we regard the system as two distinct k-out-of-n:F systems with different values of k while the system is operating under each condition. We apply the proposed model to analyze the reliability of a wireless sensor network, whose sensors switching into sleep state to save energy and shifting into listen state to detect the information alternatively. We also obtain the optimal replacement interval to redeploy the wireless sensor networks.

This paper proposes an approach for stress condition monitoring and maintenance assessment in wind turbines (WTs) through large amounts of collected data from the supervisory control and data acquisition (SCADA) system. The objectives of the proposed approach are to provide a stress condition model for health monitoring, to assess the WT’s maintenance strategies, and to provide recommendations on current maintenance schemes for future operations of the wind farm. At first, several statistical techniques, namely principal component analysis, Pearson, Spearman and Kendall correlations, mutual information, regressional ReliefF and decision trees are used and compared to assess the data for dimensionality reduction and parameter selection. Next, a normal behavior model is constructed by an artificial neural network which performs condition monitoring analysis. Then, a model based on the mathematical form of a proportional hazards model is developed where it represents the stress condition of the WT. Finally, those two models are jointly employed in order to analyze the overall performance of the WT over the study period. Several cases are analyzed with five-year SCADA data and maintenance information is utilized to develop and validate the proposed approach.

The P-S-N curves are the traditional tool for design against fatigue, which can be transformed into straight lines in the double logarithm coordinate system. There is a widely accepted assumption that the P-S-N curves will intersect at one point if they are extended to the high-stress or low-fatigue life zone. Based on this assumption and from a perspective of geometry, a new method for obtaining P-S-N curves under the condition of small sample is proposed. The proposed method makes it possible to obtain the P-S-N curves by a unified mathematical expression. The validity of the proposed method is verified by the fatigue test data of aluminum alloy 2524-T3. The result indicates that the P-S-N curves obtained by the proposed method with 17 samples agree well with those obtained by the traditional group test method with 59 samples. By comparison, the number of samples and the fatigue test time are reduced by 71.19%, while the average relative error is less than 5%.

A new approach to assess the risk associated with jetfire and poolfire accidents on an offshore oil facility using cellular automata is presented. This model simulates many accident scenarios, and the related evacuation processes, adopting a Monte Carlo approach in order to evaluate an average consequence and then a more realistic value of the risk associated with these accidents. The results of this new method are discussed and compared with the results obtained from a traditional approach over a real case study. The comparison shows that this new approach supplies lower risk-related results, still being conservative; besides, it can supply further information useful in the design phase.

This article studies a case of the banking company in Taiwan to measure the service performance by adopting reliability analysis. The service performance focuses on sending transaction data from one source (headquarters) to multiple sinks (bank branches) for a banking company transmission system. A stochastic flow network consisting of edges (transmission lines) and nodes (Internet data center or bank branches) is constructed to model such a computer system. The service-level agreement is a contract that requirements should be promised by the Internet service provider. Also, the bank needs to satisfy the customers’ requirements through the banking company transmission system. System reliability, which is defined as the probability of demand satisfaction, is studied as a key performance indicator for measuring service level of the banking company transmission system. Subsequently, sensitivity analysis is adopted to investigate the key transmission line that affects the system reliability most significantly. The system supervisor can make a decision for banking company transmission system to enhance the important edges for improving system reliability.

This article assesses the technical efficiency of different operators turning railway wheelsets on an under-floor wheel lathe. This type of lathe is a computer numerical control machine used to turn wheelsets in situ on the train. As railway wheels are turned, a certain amount of the wheel diameter is lost to restore the tread profile and full flange thickness of the wheel. The technical efficiencies of the different wheel lathe operators are assessed using a stochastic frontier analysis, while controlling for other explaining variables such as the flange thickness and the occurrence of rolling contact fatigue defects, wheel flats and cavities. Different model specifications for the stochastic frontier analysis are compared with linear mixed model specifications, showing that the stochastic frontier analysis model exhibits a better Akaike information criterion.

The fault diagnosis of hydro-turbine governing system is important to the operation of the hydropower station and the stability of the power grid. In order to improve the diagnostic accuracy and efficiency, a novel fault diagnosis method based on nonlinear output frequency response functions and a novel identification method of nonlinear output frequency response functions have been proposed and applied to the problem of hydro-turbine governing system fault diagnostics. First, the nonlinear model of hydro-turbine governing system is built. And the fault diagnosis principles based on nonlinear output frequency response functions are also introduced. Then, the disadvantages of the traditional identification method are discussed, and a novel identification method is proposed for nonlinear output frequency response functions according to the operation characteristic of hydro-turbine governing system. Finally, simulation verification and experimental studies have been presented to demonstrate the accuracy and efficiency of the proposed fault diagnosis method. The results indicate that the proposed method is simple and practical for fault diagnosis of hydro-turbine governing system.

A new general approach for obtaining system survival probability under the Poisson shock process is suggested. It takes into account the explicitly defined distribution of shock magnitude (stress) and system strength deterioration caused by the previous shocks. The approach can be used for any form of stress distribution, strength deterioration function and stress–strength interplay model. Specifically, we use the contest success function to model the stress–strength interplay and a cumulative stress ratio form function to represent strength deterioration. Numerical examples illustrating the obtained results are presented.

Besides inherent failures under normal conditions studied by classical reliability engineering, systems may also be subject to abnormal external failures, whose probabilities are hard, if not impossible, to predict, such as failures caused by natural disasters (e.g. earthquakes, floods, and tornados) and malicious attacks. Thus, abnormal external failures need to be considered when designing a system, as in reliability redundancy allocation problem. To deal with the uncertainty of abnormal external failures and based on robust optimization, this research proposes a multi-objective optimization approach that can simultaneously maximize system reliability under both normal and the worst cases of abnormal external failures for redundancy allocation problem. Since the amount of abnormal external failures is usually unknown, the worst case for each possible amount of abnormal external failures needs to be identified first. For this purpose, a new importance measure that can quantify the importance of a subset of system components with arbitrary cardinality is proposed. We also revise a widely used multi-objective evolutionary algorithm, multi-objective probabilistic solution discovery algorithm, to deal with the complexity of redundancy allocation problem. The merits of the revised algorithm are demonstrated by extensive experiments.

Variational mode decomposition is a new signal decomposition method, which can process non-linear and non-stationary signals. It can overcome the problems of mode mixing and compensate for the shortcomings in empirical mode decomposition. Permutation entropy is a method which can detect the randomness and kinetic mutation behavior of a time series. It can be considered for use in fault diagnosis. The complexity of wind power generation systems means that the randomness and kinetic mutation behavior of their vibration signals are displayed at different scales. Multi-scale permutation entropy analysis is therefore needed for such vibration signals. This research investigated a method based on variational mode decomposition and permutation entropy for the fault diagnosis of a wind turbine roller bearing. Variational mode decomposition was adopted to decompose the bearing vibration signal into its constituent components. The components containing key fault information were selected for the extraction of their permutation entropy. This entropy was used as a bearing fault characteristic value. The nearest neighbor algorithm was employed as a classifier to identify faults in a roller bearing. The experimental data showed that the proposed method can be applied to wind turbine roller bearing fault diagnosis.