Remaining Life Distribution Research Articles

An Integrated Detection-Prognostics Methodology for Components With Intermittent Faults

Abstract Some industrial components, such as valves, relay switches, and motors, occasionally experience intermittent faults (IFs) that usually disappear without any repair or intervention. This phenomenon occurs at a relatively low frequency even in components that are in an “as-good-as-new” state. However, an increase in the frequency of IFs often indicates the onset of degradation. We develop an integrated detection-prognostics model for components that exhibit IFs and whose degradation data are high-dimensional. We discuss the use of dynamic time warping (DTW) and a variational autoencoder (VAE) to perform feature engineering on the data. We then propose a hidden Markov model (HMM)-based monitoring strategy composed of two parts: (1) a detection model that tracks and flags changes in the intermittent fault frequency (IFF) and (2) a prognostic model that leverages how the transition probabilities of the HMM evolve with progressive degradation to compute the remaining life distribution (RLD) of the component. We examine the performance of our modeling framework using high-dimensional data generated from a vehicular electrical system testbed designed to accelerate the degradation of a vehicle starter motor.

Intelligent inventory management approaches for perishable pharmaceutical products in a healthcare supply chain

This study develops intelligent inventory management (IIM) approaches for managing perishable pharmaceutical products in a healthcare supply chain consisting of multiple regional hospitals and a central warehouse. The purchase price of the product is assumed to be age dependent, and the age distribution of the product is assumed to be known and is shared among all the supply chain members. Each agent representing a regional hospital or the central warehouse in the supply chain replenishes its stocks from its upstream agent. In IIM approaches, reinforcement learning methods, specifically Q-learning and Deep Q-network, are used to construct inventory policies. The IIM policies provide hospitals with near-optimal order quantities and remaining life distributions for products they order. Through many test instances, the performance of the IIM policies is compared to that of periodic review (R,s,S) policies obtained through a stochastic mixed integer programming model, solved using CPLEX and a genetic algorithm. The computational results demonstrate that the IIM policies are more cost-effective than the R,s,S policies, although the genetic algorithm has a speed advantage. Moreover, as compared with the R,s,S polices, the IIM polices have a lower possibility of product shortage, and thereby a higher service level for the patients, and a lower risk of product expiration. By implementing the IIM approaches, a healthcare system may also save storage space, in addition to having a lower total inventory cost. Sensitivity analyses are performed to derive managerial insights on the performance of the policies when the values of the key cost parameters change.

Determination of optimal post-warranty variable maintenance period for repairable second-hand product based on mean residual life

Due to the increased transactions of second-hand products in the market, the optimization of maintenance strategy for the second-hand product has become very important issue to attract a great attention from many researchers of late. This paper proposes a new post-warranty strategy with a variable self-maintenance period for the second-hand product, assuming that the product is replaced by another one on the first failure following a fixed length of post-warranty self-maintenance period. During the non-renewing warranty period, the product is subject to preventive maintenance periodically at a prorated cost while only minimal repair is implemented at each failure by the dealer. The main goal of this study is to determine an optimal length of post-warranty self-maintenance period which minimizes the expected cost rate per unit time during the product’s life cycle from the user’s perspective. This approach considers not only the periodic preventive maintenance during the warranty period, but also the remaining life distribution of the product after the warranty expires, which is the significant difference of this work from many existing maintenance policies. For this purpose, we formulate the expected length of life cycle and evaluate the expected total cost incurred during the life cycle of the second-had product which is purchased at the age of [Formula: see text] The existence of the optimal self-maintenance period is proved analytically under mild conditions and the proposed maintenance model is compared with an existing model with regard to the expected cost rate. Finally, assuming that the life distribution of the product follows a Weibull distribution, the effect of relevant parameters on the optimal self-maintenance period is analyzed numerically.

An integrated optimization framework for multi-component predictive analytics in wind farm operations & maintenance

Recent years have seen an unprecedented growth in the use of sensor data to guide wind farm operations and maintenance. Emerging sensor-driven approaches typically focus on optimal maintenance procedures for single turbine systems, or model multiple turbines in wind farms as single component entities. In reality, turbines are composed of multiple components that dynamically interact throughout their lifetime. These interactions are central for realistic assessment and control of turbine failure risks. In this paper, an integrated framework that combines i) real-time degradation models used for predicting remaining life distribution of each component, with ii) mixed integer optimization models and solution algorithms used for identifying optimal wind farm maintenance and operations is proposed. Maintenance decisions identify optimal times to repair every component, which in turn, determine the failure risk of the turbines. More specifically, optimization models that characterize a turbine's failure time as the first time that one of its constituent components fail - a systems reliability concept called competing risk is developed. The resulting turbine failures impact the optimization of wind farm operations and revenue. Extensive experiments conducted for multiple wind farms with 300 wind turbines - 1200 components - showcases the performance of the proposed framework over conventional methods.

Residual Lifetime Prediction with Multistage Stochastic Degradation for Equipment

Residual useful lifetime (RUL) prediction plays a key role of failure prediction and health management (PHM) in equipment. Aiming at the problems of residual life prediction without comprehensively considering multistage and individual differences in equipment performance degradation at present, we explore a prediction model that can fit the multistage random performance degradation. Degradation modeling is based on the random Wiener process. Moreover, according to the degradation monitoring data of the same batch of equipment, we apply the expectation maximization (EM) algorithm to estimate the prior distribution of the model. The real-time remaining life distribution of the equipment is acquired by merging prior information of real-time degradation data and historical degradation monitoring data. The accuracy of the proposed model is demonstrated by analyzing a practical case of metalized film capacitors, and the result shows that a better RUL estimation accuracy can be provided by our model compared with existing models.

GAN 기반 데이터생성을 통한 가스배관의 잔여수명분포 추정

GAN 기반 데이터생성을 통한 가스배관의 잔여수명분포 추정

Statistical modeling of the residual life of an oil and gas pipeline with axial crack-like defects

According to the deterministic formula using the statistical modeling method, the remaining life distribution and indices of the oil and gas pipeline reliability are found when the characteristics of the strength, stress and the dimensions of the defects in the axial and radial directions are subject to the normal law. As a result, the safe service life limits of the diagnosed oil and gas pipelines were established.

Real-Time Estimation of Mean Remaining Life Using Sensor-Based Degradation Models

Advances in sensor technology have led to an increased interest in using degradation-based sensory information to predict the remaining lives of partially degraded components and systems. This paper presents a stochastic degradation modeling framework for computing and continuously updating remaining life distributions (RLDs) using in situ degradation signals acquired from individual components during their operational lives. Unfortunately, these sensory-updated RLDs cannot be characterized using parametric distributions and their moments do not exist. Such difficulties hinder the implementation of this sensor-based framework, especially from the standpoint of computational efficiency of embedded algorithms. In this paper, we identify an approximate procedure by which we can compute a conservative mean of the sensory-updated RLDs and express the mean and variance using closed-form expressions that are easy to evaluate. To accomplish this, we use the first passage time of Brownian motion with positive drift, which follows an inverse Gaussian distribution, as an approximation of the remaining life. We then show that the mean of the inverse Gaussian is a conservative lower bound of the mean remaining life using Jensen’s inequality. The results are validated using real-world vibration-based degradation information.

Sensor-driven prognostic models for equipment replacement and spare parts inventory

Accurate predictions of equipment failure times are necessary to improve replacement and spare parts inventory decisions. Most of the existing decision models focus on using population-specific reliability characteristics, such as failure time distributions, to develop decision-making strategies. Since these distributions are unaffected by the underlying physical degradation processes, they do not distinguish between the different degradation characteristics of individual components of the population. This results in less accurate failure predictability and hence less accurate replacement and inventory decisions. In this paper, we develop a sensor-driven decision model for component replacement and spare parts inventory. We integrate a degradation modeling framework for computing remaining life distributions using condition-based in situ sensor data with existing replacement and inventory decision models. This enables the dynamic updating of replacement and inventory decisions based on the physical condition of the equipment.

Sensory-Updated Residual Life Distributions for Components With Exponential Degradation Patterns

Research on interpreting data communicated by smart sensors and distributed sensor networks, and utilizing these data streams in making critical decisions stands to provide significant advancements across a wide range of application domains such as maintenance management. In this paper, a stochastic degradation modeling framework is developed for computing and continuously updating residual life distributions of partially degraded components. The proposed degradation methodology combines population-specific degradation characteristics with component-specific sensory data acquired through condition monitoring in order to compute and continuously update remaining life distributions of partially degraded components. Two sensory updating procedures are developed and validated using real-world vibration-based degradation information acquired from rolling element thrust bearings. The results are compared with two benchmark policies and illustrate the benefits of the sensory updated degradation models proposed in this paper. Note for Practitioners-The proposed degradation-based prognostic methodology provides a comprehensive assessment of the current and future degradation states of partially degraded components by combining population-specific degradation or reliability information with real-time sensory health monitoring data. It is specifically beneficial for cases where degradation occurs in a cumulative manner and the degradation signal can be approximated by an exponential functional form. To implement this methodology, it is necessary: 1) to identify the physical phenomena associated with the evolution of the degradation process (spalling and wear herein); 2) choose the appropriate condition monitoring technology to monitor this phenomena (accelerometers); 3) identify a characteristic pattern in the sensory information to help develop a degradation signal (exponential growth); and 4) identify a failure threshold associated with the degradation signal. The first step in implementing this prognostic methodology is to obtain prior information related to stochastic parameters f the exponential model. This may require fitting some sample degradation signals with an exponential functional form and noting the values of the exponential parameters, or using subjective prior distributions. The second step is to acquire sensory information and begin updating the prior distribution. The updating frequency will dictate which expressions are used to compute the posterior distributions. Once the posterior means, variances, and correlation are computed, the truncated CDF of the residual life can be evaluated using (10) and (11). Note that the truncation is necessary to preclude negative values of the remaining life. Practitioners can implement this methodology using a simple spreadsheet. Since the residual life distributions are skewed, it is reasonable to utilize the median as a measure of the central tendency and, hence, an alternative estimate for the expected value of the remaining life

Nonparametric test for monotone variance residual life class of life distributions with hypothesis testing applications

The problem of testing exponentiality against decreasing (increasing) variance remaining life distribution DVRL (IVRL) is investigated. For this aging properties, a nonparametric procedure ( U-Test) is presented. The percentiles of this test are tabulated for sample sizes n=5(1)40. It is shown that the proposed test is simple, has high relative efficiency for some commonly used alternatives and enjoys a good power. Data of 40 patients suffering from blood cancer disease (Leukemia) is considered as a practical application of the proposed test in the medical sciences.

On some aspects of variance remaining life distributions

Some new variance remaining properties of life distributions are introduced and studied. Their relationships with the mean remaining life distributions are established. The survival function of a component is expressed in terms of its variance remaining life. The distribution with constant variance remaining life is shown to be exponential. The variance remaining life of some known life distributions in reliability namely gamma and Weibull are calculated.

A comparison of lumped-parameter and diffusional models describing the effects of the outlet boundary conditions on the mixing in flow systems

A comparison is presented between two methods of modeling the dispersion and mixing processes in large (geophysical) turbulent flow systems, such as estuaries, coastal waters, and atmospheric systems. One method is based on the diffusional model and the other is a lumped-parameter model, referred to as “box-mode” or compartmental model. The models are compared on the basis of model free parameters defined here: the purging rate and remaining life distribution. It is found that there exists a good agreement between the results of both models for a one-dimensional system. The problem of the boundary condition at the outlet is discussed; it is concluded that the use of a mass transfer coefficient may be justified in some cases where inflows at the outlet (e.g. due to tidal motion) can be estimated. When these inflows are strong a sink at the outlet can be taken as a boundary condition.