Bayesian planning of step‐stress accelerated degradation tests under various optimality criteria

Step‐stress accelerated degradation testing (SSADT) has become a common approach to predicting lifetime for highly reliable products that are unlikely to fail in a reasonable time under use conditions or even elevated stress conditions. In literature, the planning of SSADT has been widely investigated for stochastic degradation processes, such as Wiener processes and gamma processes. In this paper, we model the optimal SSADT planning problem from a Bayesian perspective and optimize test plans by determining both stress levels and the allocation of inspections. Large‐sample approximation is used to derive the asymptotic Bayesian utility functions under 3 planning criteria. A revisited LED lamp example is presented to illustrate our method. The comparison with optimal plans from previous studies demonstrates the necessity of considering the stress levels and inspection allocations simultaneously.     

Economic design of an attribute control chart for monitoring mean life based on multiple deferred state sampling

In this paper, we design an attribute np control chart using multiple deferred state (MDS) sampling under Weibull distribution based on time truncated life test. This chart is constructed for monitoring the variation of mean life of the product in a manufacturing process. The optimal parameters of MDS sampling and the control limit coefficients are determined so that the in‐control average run length (ARL) is as close as to the target ARL. The optimal parameters of MDS sampling are sample size and number of successive subgroups required for declaring the current state of process. Out‐of‐control ARL is considered as a measure of the performance of proposed chart and reported with determined optimal parameters for various shift constants. The out‐of‐control ARL of the proposed chart obtained under various distributions is compared with each other. The performance of proposed control chart is compared with the performance of the existing control chart designed under single sampling. In addition, the economic design of proposed chart using variable sampling interval scheme is discussed, and sensitivity analysis on expected costs is also investigated.     

基于广义逆理论的步进应力加速寿命试验的最优设计

步进应力加速寿命试验的最优设计问题早在20世纪60年代就已提出.本文在应力个数和未知参数个数不相等的情况下,以MLE渐近方差最小为准则,运用广义加号逆的理论,解决指数分布时定数截尾寿命试验和定时截尾寿命试验的最优设计问题.通过理论推导,得出定数截尾寿命试验时的不同应力下失效产品数的安排,以及定时截尾寿命试验时的最优应力变换点.并且,对每一种情况进行了数据模拟,结果显示,这些理论和方法在实际运用中是可行的和有效的.     

Bayesian optimal life‐testing plan under the balanced two sample type‐II progressive censoring scheme

Joint progressive censoring schemes are quite useful to conduct comparative life‐testing experiment of different competing products. Recently, Mondal and Kundu (“A New Two Sample Type‐II Progressive Censoring Scheme,” Commun Stat‐Theory Methods; 2018) introduced a joint progressive censoring scheme on two samples known as the balanced joint progressive censoring (BJPC) scheme. Optimal planning of such progressive censoring scheme is an important issue to the experimenter. This article considers optimal life‐testing plan under the BJPC scheme using the Bayesian precision and D‐optimality criteria, assuming that the lifetimes follow Weibull distribution. In order to obtain the optimal BJPC life‐testing plans, one needs to carry out an exhaustive search within the set of all admissible plans under the BJPC scheme. However, for large sample size, determination of the optimal life‐testing plan is difficult by exhaustive search technique. A metaheuristic algorithm based on the variable neighborhood search method is employed for computation of the optimal life‐testing plan. Optimal plans are provided under different scenarios. The optimal plans depend upon the values of the hyperparameters of the prior distribution. The effect of different prior information on optimal scheme is studied.     

Models and algorithms for a staff scheduling problem

We present mathematical models and solution algorithms for a family of staff scheduling problems arising in real life applications. In these problems, the daily assignments to be performed are given and the durations (in days) of the working and rest periods for each employee in the planning horizon are specified in advance, whereas the sequence in which these working and rest periods occur, as well as the daily assignment for each working period, have to be determined. The main objective is the minimization of the number of employees needed to perform all daily assignments in the horizon. We decompose the problem into two steps: the definition of the sequence of working and rest periods (called pattern) for each employee, and the definition of the daily assignment to be performed in each working period by each employee. The first step is formulated as a covering problem for which we present alternative ILP models and exact enumerative algorithms based on these models. Practical experience shows that the best approach is based on the model in which variables are associated with feasible patterns and generated either by dynamic programming or by solving another ILP. The second step is stated as a feasibility problem solved heuristically through a sequence of transportation problems. Although in general this procedure may not find a solution (even if one exists), we present sufficient conditions under which our approach is guaranteed to succeed. We also propose an iterative heuristic algorithm to handle the case in which no feasible solution is found in the second step. We present computational results on real life instances associated with an emergency call center. The proposed approach is able to determine the optimal solution of instances involving up to several hundred employees and a working period of up to 6 months. Mathematics Subject Classification (2000): 90B70, 90C10, 90C27, 90C39, 90C57, 90C59     

指数分布场合下双应力步加试验的设计

本文利用指数分布场合下双应力无交互作用时定时步进应力加速寿命试验~(简称步加试验) 的非退化试验设计与退化试验设计的关系,将双应力试验设计转化成单应力试验设计, 在``使得工作应力下产品对数分位寿命的极大似然估计的渐近方差最小'的最优准则下, 给出双应力无交互作用时定时步加试验的设计方法, 同时证明该设计也是D-最优的.     

Bayesian planning of optimal step-stress accelerated life test for log-location-scale distributions

This paper introduces some Bayesian optimal design methods for step-stress accelerated life test planning with one accelerating variable, when the acceleration model is linear in the accelerated variable or its function, based on censored data from a log-location-scale distributions. In order to find the optimal plan, we propose different Monte Carlo simulation algorithms for different Bayesian optimal criteria. We present an example using the lognormal life distribution with Type-I censoring to illustrate the different Bayesian methods and to examine the effects of the prior distribution and sample size. By comparing the different Bayesian methods we suggest that when the data have large(small) sample size B₁(τ) (B₂(τ)) method is adopted. Finally, the Bayesian optimal plans are compared with the plan obtained by maximum likelihood method.     

Optimal Robust Design of Step Stress Accelerated Life Test Scheme under Weibull Distribution

The traditional accelerated life test scheme is necessary to give the rough values of some model parameters in advance, but the influence of fluctuation on the stability of test scheme is irregulared. Based on the prior life test information, this paper aims to minimize the mean and variance of asymptotic variance of $p$-quantile life estimate under normal test stress level, using maximum likelihood estimation theory and Nelson cumulative failure principle, the optimal robust design mathematical model of step stress accelerated life test scheme with uncertainty parameters under Weibull distribution is established. The results of optimal robust design of step stress accelerated life test scheme for electrical connectors show that: comparing with the optimal design of step stress test scheme in the literature, the optimal robust design scheme is not sensitive to the uncertainty of model parameters when the asymptotic variance of median life estimate is basically the same; Comparing with the optimal design of constant accelerated life test scheme, when the statistical accuracy of test data is basically the same, the number of samples required can be reduced by 1/5, and the test time can be reduced by about 1/4.     

Bayesian accelerated life test plans for series systems with Weibull component lifetimes

This article presents optimal Bayesian accelerated life test plans for series systems under Type-I censoring scheme. First, the component lifetimes are assumed to follow independent Weibull distributions. The scale parameters of Weibull lifetime distributions are related to the external stress variable through a general stress translation function. For a fixed number of design points, optimal Bayesian ALT plans are first obtained by solving constrained optimization problems under two different Bayesian design criteria. The global optimality of the resulting fixed-point optimal designs is then verified via the General Equivalence Theorem. This article also provides the optimized compromise ALT plans which are extremely useful in real-life applications. A detailed sensitivity analysis is then performed to find out the effect of various planning inputs on the resulting optimal Bayesian ALT plans. A simulation study is then conducted to visualize the resulting sampling variations from the optimal Bayesian ALT plans. Finally, this article considers a series system with dependent component lifetimes. Optimal ALT plans are obtained assuming a Gamma frailty model.     

Robust designs for Haar wavelet approximation models

In this paper, we discuss the construction of robust designs for heteroscedastic wavelet regression models when the assumed models are possibly contaminated over two different neighbourhoods: G ₁ and G ₂ . Our main findings are: (1) A recursive formula for constructing D‐optimal designs under G ₁ ; (2) Equivalency of Q‐optimal and A‐optimal designs under both G ₁ and G ₂ ; (3) D‐optimal robust designs under G ₂ ; and (4) Analytic forms for A‐ and Q‐optimal robust design densities under G ₂ . Several examples are given for the comparison, and the results demonstrate that our designs are efficient. Copyright © 2010 John Wiley & Sons, Ltd.     

What is an infinite design?

It is usually assumed that an infinite design is a design with infinitely many points. This encompasses a myriad of structures, some nice and others not. In this paper we consider examples of structures that we would not like to call designs, and investigate additional conditions that exclude such anomalous structures. In particular, we expect a design to be regular, the complement of a design to be a design, and a t‐design to be an s‐design, for all 0 < s ≤ t. These are all properties that can be taken for granted with finite designs, and for infinite Steiner systems. We present a new definition of an infinite t‐design, and give examples of structures that satisfy this definition. We note that infinite designs considered in the literature to date satisfy our definition. We show that infinite design theory does not always mirror finite design theory, for example there are examples of designs with υ > b. © 2002 Wiley Periodicals, Inc. J Combin Designs 10: 79–91, 2002; DOI 10.1002/jcd.10005     

A technique for constructing non‐embeddable quasi‐residual designs

We propose a technique for constructing two infinite families of non‐embeddable quasi‐residual designs as soon as one such design satisfying certain conditions exists. The main tools are generalized Hadamard matrices and balanced generalized weighing matrices. Starting with a specific non‐embeddable quasi‐residual 2‐(27,9,4) design, we construct for every positive integer m a non‐embeddable 2‐(3^m,3^m?1,(3^m?1?1)/2)‐design, and, if r_m=(3^m?1)/2 is a prime power, we construct for every positive integer n a non‐embeddable design. For each design in these families, a symmetric design with the corresponding parameters is known to exist. © 2002 Wiley Periodicals, Inc. J Combin Designs 10: 160–172, 2002; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/jcd.900     

Adding more observations to saturated D‐optimal resolution III two‐level factorial designs

We consider the class of saturated main effect plans for the 2^k factorial. With these saturated designs, the overall mean and all main effects can be unbiasedly estimated provided that there are no interactions. However, there is no way to estimate the error variance with such designs. Because of this and other reasons, we like to add some additional runs to the set of (k+1) runs in the D‐optimal design in this class. Our goals here are: (1) to search for s additional runs so that the resulting design based on (k+s+1) runs yields a D‐optimal design in the class of augmented designs; (2) to classify all the runs into equivalent classes so that the runs in the same equivalent class give us the same value of the determinant of the information matrix. This allows us to trade runs for runs if this becomes necessary; (3) to obtain upper bounds for determinant of the information matrices of augmented designs. In this article we shall address these approaches and present some new results. © 2002 Wiley Periodicals, Inc. J Combin Designs 11: 51–77, 2003; Published online in Wiley InterScience ( www.interscience.wiley.com ). DOI 10.1002/jcd.10026     

逐步区间删失Weibull分布的Bayesian稳健设计

逐步区间删失是获取高可靠性产品相关信息的一种重要方法.文章研究了产品寿命服从Weibull分布,带有随机移除逐步区间删失寿命试验的最优设计问题.采用极大似然方法获取模型参数的估计及其信息矩阵.利用Bayesian方法处理模型参数未知情况下设计准则对模型参数的依赖问题,获得了模型参数估计的Bayesian稳健设计准则.在考虑试验费用有限制的条件下,给出了获得最优稳健设计非线性混合整数算法.同时对先验选取及约束参数设定的敏感性做了分析.数值结果表明文章提出的方法是有效可行的.     

A class of group divisible 3‐designs and their applications

In this article, we first show that a group divisible 3‐design with block sizes from {4, 6}, index unity and group‐type 2^m exists for every integer m≥ 4 with the exception of m = 5. Such group divisible 3‐designs play an important role in our subsequent complete solution to the existence problem for directed H‐designs DH_λ(m, r, 4, 3)s. We also consider a way to construct optimal codes capable of correcting one deletion or insertion using the directed H‐designs. In this way, the optimal single‐deletion/insertion‐correcting codes of length 4 can be constructed for all even alphabet sizes. © 2008 Wiley Periodicals, Inc. J Combin Designs 17: 136–146, 2009     

Single and multi‐interval Legendre spectral methods in time for parabolic equations

In this article, we take the parabolic equation with Dirichlet boundary conditions as a model to present the Legendre spectral methods both in spatial and in time. Error analysis for the single/multi‐interval schemes in time is given. For the single interval spectral method in time, we obtain a better error estimate in L²‐norm. For the multi‐interval spectral method in time, we obtain the L²‐optimal error estimate in spatial. By choosing approximate trial and test functions, the methods result in algebraic systems with sparse forms. A parallel algorithm is constructed for the multi‐interval scheme in time. Numerical results show the efficiency of the methods. The methods are also applied to parabolic equations with Neumann boundary conditions, Robin boundary conditions and some nonlinear PDEs. © 2005 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq, 2006     

A Simple Approach for Multi‐fidelity Experimentation Applied to Financial Engineering

In practical applications, information about the accuracy or ‘fidelity’ of alternative surrogate systems may be ambiguous and difficult to determine. To address this problem, we propose to treat surrogate system fidelity level as a categorical factor in optimal response surface design. To design the associated experiments, we apply the Expected Integrated Mean Squared Error optimal design criterion, which takes into account both variance and bias errors. The performance of the proposed design was compared using three test cases to four types of alternatives using the Empirical Integrated Squared Error. Because of its ability to foster relatively accurate predictions, the proposed design is recommended in fidelity experimental design, particularly when the experimenters lack sufficient information about the fidelity levels of surrogate systems. The method was applied to the case of intraday trading optimization in which data were collected from the Taiwan Futures Exchange. We also calculated the implied volatility from the Merton's Jump‐diffusion model via the fast Fourier transform algorithm with three different models of varying fidelity levels. Copyright © 2014 John Wiley & Sons, Ltd.     

Common intersection designs

Motivated by an application to sensor networks, Lee and Stinson [ 6 ] defined a new type of set system termed a “common intersection design.” Briefly, a µ‐common intersection design is a 1‐design in which no pair of points occurs in more than one block, and in which any two disjoint blocks intersect at least µ blocks in common. In general, we want to maximize µ as a function of the other parmameters of the design. In this paper, we analyze combinatorial properties of common intersection designs. We determine necessary conditions for “optimal” common intersection designs and provide several existence results. Connections with other types of designs are pointed out. © 2005 Wiley Periodicals, Inc. J Combin Designs 14: 251–269, 2006     

Comparing methods for design follow‐up: revisiting a metal‐cutting case study

Adding another fraction to an initial fractional factorial design is often required to resolve ambiguities with respect to aliasing of factorial effects from the initial experiment and/or to improve estimation precision. Multiple techniques for design follow‐up exist; the choice of which is often made on the basis of the initial design and its analysis, resources available, experimental objectives, and so on. In this paper, we compare four design follow‐up strategies: foldover, semifoldover, D‐optimal, and Bayesian (MD‐optimal) in the context of a metal‐cutting case study previously utilized to compare fractional factorials of different run sizes. Follow‐up designs are compared for each of a , , and Plackett–Burman initial experiments. Our empirical results suggest that a single follow‐up strategy does not outperform all others in every situation. This case study serves to illustrate design augmentation possibilities for practitioners and provides some basis for the selection of a follow‐up experiment. Copyright © 2013 John Wiley & Sons, Ltd.     

Probabilistic analysis of power assignments

A fundamental problem for wireless ad hoc networks is the assignment of suitable transmission powers to the wireless devices such that the resulting communication graph is connected. The goal is to minimize the total transmit power in order to maximize the life‐time of the network. Our aim is a probabilistic analysis of this power assignment problem. We prove complete convergence for arbitrary combinations of the dimension d and the distance‐power gradient p. Furthermore, we prove that the expected approximation ratio of the simple spanning tree heuristic is strictly less than its worst‐case ratio of 2. Our main technical novelties are two‐fold: First, we find a way to deal with the unbounded degree that the communication network induced by the optimal power assignment can have. Minimum spanning trees and traveling salesman tours, for which strong concentration results are known in Euclidean space, have bounded degree, which is heavily exploited in their analysis. Second, we apply a recent generalization of Azuma‐Hoeffding's inequality to prove complete convergence for the case for both power assignments and minimum spanning trees (MSTs). As far as we are aware, complete convergence for p > d has not been proved yet for any Euclidean functional. © 2017 The Authors Random Structures & Algorithms Published by Wiley Periodicals, Inc., 51, 483–505, 2017