An integrated model of statistical process control and condition-based maintenance for deteriorating systems

This paper proposes an integrated model of statistical process control and condition-based maintenance for a deteriorating system. We study a system that will not be as good as new after a preventive maintenance and can only survive a certain number of preventive maintenances. The system is modeled as a geometric process and monitored by an \(\bar{X}\) control chart. By analyzing the evolution of the system in different scenarios, we establish a mathematical model to minimize the expected cost during the expected cycle time that can be used to make an optimal replacement policy in applications. A computational scheme is presented and illustrated through a numerical example. A sensitivity analysis is performed to investigate the effect of statistical constraint, mean shift, and the parameters of the system.     

Ontology concepts and tools for statistical genomics

In computer science, an ontology is any formally structured vocabulary covering a conceptual domain. Gene Ontology (GO) is a structured collection of terms defining biological processes, cellular components, or molecular functions for the purpose of characterizing gene products and functions. The structure of GO is a directed acyclic graph (DAG) with typed edges. We describe a simple formalism for working with ontologies for statistical purposes, and define object-ontology complexes, which encode the usage of the vocabulary to label objects under analysis. Recently developed concepts of information content and semantic similarity are evaluated and used to explore the association between LocusLink loci and GO. We investigate relations between GO DAG structure, association evidence codes and term information content, illustrate computation of semantic similarities of genes within and between clusters discovered in a microarray, and describe a more general ontology and its use in inference on genetic network structure.     

Nilpotent block designs I: Basic concepts for Steiner triple and quadruple systems

This paper discusses the concepts of nilpotence and the center for Steiner Triple and Quadruple Systems. The discussion is couched in the language of block designs rather than algebras. Nilpotence is closely connected to the well known doubling and tripling constructions for these designs. A sample result: a point p in an STS is projective if every triangle containing p generates the 7-element Fano plane; the p-center of the STS is the set of all projective points and is a projective geometry over GF(2). © 1999 John Wiley & Sons, Inc. J Combin Designs 7: 157–171, 1999     

User-friendly statistical concepts for process monitoring

Many standard statistical process control techniques involve sophisticated mathematical concepts, which are frequently misunderstood and misused by their users. This means, as this paper argues, that the techniques, and the terminology and concepts underlying them, are inappropriate for their intended uses and users. The paper considers the areas of the statistical background which cause difficulties. It goes on to describe some alternative concepts, techniques and terminology, which are likely to be conceptually simpler and more ‘user-friendly’ (and in some cases more accurate and robust). None of the ideas require users to be familiar with the standard deviation. We also suggest that the approach of reengineering the conceptual background to suit the context of users and uses may be appropriate to other areas of practical statistics and OR. The paper identifies some general principles for achieving this.     

Logic and probabilistic systems

Following some ideas of Roberto Magari, we propose trial and error probabilistic functions, i.e. probability measures on the sentences of arithmetic that evolve in time by trial and error. The set of the sentences that get limit probability 1 is a theory, in fact can be a complete set. We prove incompleteness results for this setting, by showing for instance that for every there are true sentences that get limit probability less than . No set as above can contain the set of all true sentences, although we exhibit some containing all the true sentences. We also consider an approach based on the notions of inner probability and outer probability, and we compare this approach with the one based on trial and error probabilistic functions. Although the two approaches are shown to be different, we single out an important case in which they are equivalent. Received March 20, 1995     

Integral and probabilistic representations for systems of elliptic equations

We present probabilistic representations for some systems of elliptic equations constructed as expectations of functionals of some specific Markov chains, in particular, the walk on spheres processes. These representations are deduced from converse mean value theorems that we prove for the equations under analysis, especially the Lamé equation of elasticity theory. We construct Monte Carlo algorithms, and estimate the variances and the cost.     

Versatile thinking and the learning of statistical concepts

Statistics was for a long time a domain where calculation dominated to the detriment of statistical thinking. In recent years, the latter concept has come much more to the fore, and is now being both researched and promoted in school and tertiary courses. In this study, we consider the application of the concept of flexible or versatile thinking to statistical inference, as a key attribute of statistical thinking. Whilst this versatility comprises process/object, visuo/analytic and representational versatility, we concentrate here on the last aspect, which includes the ability to work within a representation system (or semiotic register) and to transform seamlessly between the systems for given concepts, as well as to engage in procedural and conceptual interactions with specific representations. To exemplify the theoretical ideas, we consider two examples based on the concepts of relative comparison and sampling variability as cases where representational versatility may be crucial to understanding. We outline the qualitative thinking involved in representations of relative density and sample and population distributions, including mathematical models and their precursor, diagrammatic forms.     

Semi-Markov modulated Poisson process: probabilistic and statistical analysis

We consider a Poisson process that is modulated in such a way that the arrival rate at any time depends on the state of a semi-Markov process. This presents an interesting generalization of Poisson processes with important implications in real life applications. Our analysis concentrates on the transient as well as the long term behaviour of the arrival count and the arrival time processes. We discuss probabilistic as well as statistical issues related to various quantities of interest.     

Preventive maintenance for homogeneous and heterogeneous systems

We consider optimal preventive maintenance for homogeneous and heterogeneous systems with major (critical) and minor (noncritical) failures. A major failure results in a replacement of a failed system, whereas minor failures can be minimally instantaneously repaired. Distinct from the homogeneous case, where the process of minimal repairs is the Poisson process, the process of minimal repairs in the heterogeneous case is the mixed Poisson process that does not possess the memoryless property. This enables considering the number of minimal repairs as the decision parameter for the corresponding optimal preventive maintenance policy. The proposed approach is theoretically justified, and the detailed illustrative numerical examples are presented.     

Optimization of spare parts for hierarchicall decomposable systems

Complex technical systems are often hierarchically composed of exchangeable modules, which themselves may be composed of exchangeable submodules. The hierarchy of modules of such systems can be represented by trees. Defective systems are repaired in a supply depot system by exchanging the defective modules from a stock of spare parts. The exchanged defective modules are repaired for resupply. In this paper, a dynamic programming algorithm for optimizing the stocks of spare parts, under a budget constraint, is presented. The optimization criterion is the minimization of the backlog of entire systems at the supply depot system. An implementation of the algorithm is described and some illustrative examples are given.     

ON THE STATISTICAL MECHANICS OF PROBABILISTIC MODELS OF NEURAL NETWORKS

ＯＮＴＨＥＳＴＡＴＩＳＴＩＣＡＬＭＥＣＨＡＮＩＣＳＯＦＰＲＯＢＡＢＩＬＩＳＴＩＣ　ＭＯＤＥＬＳＯＦＮＥＵＲＡＬＮＥＴＷＯＲＫＳＱＩＡＮＭＩＮＰＩＮＧ（钱敏平）（ＤｅｐａｒｔｍｅｎｔｏｆＰｒｏｂａｂｉｌｉｔｙａｎｄＳｔａｔｉｓｔｉｃｓ，ＢｅｉｊｉｎｇＵ...     

A stochastic model for joint spare parts inventory and planned maintenance optimisation

Spare parts demands are usually generated by the need of maintenance either preventively or at failures. These demands are difficult to predict based on historical data of past spare parts usages, and therefore, the optimal inventory control policy may be also difficult to obtain. However, it is well known that maintenance costs are related to the availability of spare parts and the penalty cost of unavailable spare parts consists of usually the cost of, for example, extended downtime for waiting the spare parts and the emergency expedition cost for acquiring the spare parts. On the other hand, proper planned maintenance intervention can reduce the number of failures and associated costs but its performance also depends on the availability of spare parts. This paper presents the joint optimisation for both the inventory control of the spare parts and the Preventive Maintenance (PM) inspection interval. The decision variables are the order interval, PM interval and order quantity. Because of the random nature of plant failures, stochastic cost models for spare parts inventory and maintenance are derived and an enumeration algorithm with stochastic dynamic programming is employed for finding the joint optimal solutions over a finite time horizon. The delay-time concept developed for inspection modelling is used to construct the probabilities of the number of failures and the number of the defective items identified at a PM epoch, which has not been used in this type of problems before. The inventory model follows a periodic review policy but with the demand governed by the need for spare parts due to maintenance. We demonstrate the developed model using a numerical example.     

Failure replacement and preventive maintenance spare parts ordering policy

This paper addresses inventory policy for spare parts, when demand for the spare parts arises due to regularly scheduled preventive maintenance, as well as random failure of units in service. A stochastic dynamic programming model is used to characterize an ordering policy which addresses both sources of demand in a unified manner. The optimal policy has the form (s(k),S(k)), where k is the number of periods until the next scheduled preventive maintenance operation. The nature of the (s(k),S(k)) policy is characterized through numeric evaluation. The efficiency of the optimal policy is evaluated, relative to a simpler policy which addresses the failure replacement and preventive maintenance demands with separate ordering policies.     

A framework for the probabilistic analysis of hierarchical planning systems

As we have argued in previous papers, multi-level decision problems can often be modeled as multi-stage stochastic programs, and hierarchical planning systems designed for their solution, when viewed as stochastic programming heuristics, can be subjected to analytical performance evaluation. The present paper gives a general formulation of such stochastic programs and provides a framework for the design and analysis of heuristics for their solution. The various ways to measure the performance of such heuristics are reviewed, and some relations between these measures are derived. Our concepts are illustrated on a simple two-level planning problem of a general nature and on a more complicated two-level scheduling problem.     

On the use of zadeh's probabilistic definition for testing statistical hypotheses from fuzzy information

A statistical hypothesis is an assertion about the distribution of an experiment. We consider the study of the problem of testing a statistical hypothesis (that is, the problem of concluding whether or not the hypothesis is correct) on the basis of data from the experiment, when its outcomes do not provide exact but rather fuzzy information.For establishing optimality criteria of testing we will use the definition of probability of a fuzzy event, given by Zadeh, in order to extend both Neyman-Pearson and Bayes theories, to the fuzzy framework.Then, we will analyze several properties for the new criteria. Particularly, the goodness of optimal procedures in both the fuzzy and the nonfuzzy situation, will be compared for each criterion.Finally, we will apply the extended criteria for testing simple hypotheses. This application leads us to prefer Bayesian procedures to Neyman-Pearson procedures in the fuzzy context.     

Basic concepts for a theory of evaluation: The aggregative operator

Starting with an explication of the “aggregative”-concept and deducing a general structure which satisfies a number of minimal requirements (properties of clustering) the main features of a new mathematical theory — called “theory of evaluation” — are developed. The theory sheds new light on such well-known concepts as membership, conjunction and disjunction and seems to be a very promising tool to handle representation problems as they grow from the fields of theory of fuzzy set, and its many applications, of human decision making and of multicriteria analysis.     

Modelling preventive maintenance for deteriorating repairable systems

This paper is concerned with practical methods for the analysisand modelling of data for repairable systems which are subjectto preventive maintenance (PM) and still have an increasingrate of occurrence of failures. Aspects of testing for trendand fitting a nonhomogeneous Poission process to data are discussedModels for scheduling preventive maintenance to minimize costor maximize availability are proposed. They show that the optimalPM cycle interval for these systems decreases with increasingequipment age. One-cycle and two-cycle finite-time-zone replacementmodels are also developed to decide the optimal time for replacingthe equipment in current use.     

On unified concepts of detectability and observability for continuous-time stochastic systems

This paper is concerned with detectability and observability of continuous-time stochastic linear systems. By adopting the idea used in defining these two concepts for time-varying systems and Markovian jump linear systems that have been studied in the literature, corresponding definitions for continuous-time stochastic linear systems are proposed. These two definitions are not only able to unify some recent definitions on these two concept reported in the literature, but also allow us to propose an efficient rank criterion to test observability of continuous-time stochastic linear systems. It seems that this rank criterion is quite analogous to the rank criterion for deterministic linear systems. With the help of these two concepts and the new criteria, the stochastic Lyapunov equation is revisited and some recent published work on this equation are generalized. Numerical examples are given to illustrate the effectiveness of the proposed approach.     

An optimal maintenance policy for skipping imminent preventive maintenance for systems experiencing random failures

In this study we investigate systems that experience random failures and establish decision rules for performing renewal maintenance; that is, a preventive replacement (PR) policy. We seek a policy that is both simple to execute from the point of view of the maintenance planner but also a policy that is an improvement on existing schemes. We show that our policy is a hybrid of traditional time-based and age-based schemes and one that yields considerable cost savings. Our hybrid policy involves two decision variables. One decision variable is the time between PRs. Hence, for the maintenance planner, the times at which PRs are performed are chronologically fixed. Random failures can occur, however, and the machine receives an emergency renewal (ER) at these times. Hence, within these chronological times, a second decision time is identified. Should an ER occur between the start of a cycle and this second decision time, then the planned PR would still be performed at the end of the cycle. However, if the first ER occurs after this second decision time, then the PR at the end of the cycle is skipped over and the next planned PR would take place at the end of the subsequent cycle. With this simple mechanism, PRs that follow on too closely after an ER are avoided, thus saving the unnecessary expense. Numerical examples are given to examine the validity of the model, using four different failure density functions, namely Weibull, normal, uniform, and negative exponential.     

Statistics on the table: The history of statistical concepts and methods

Feel like writing a review for The Mathematical Intelligencer? You are welcome to submit an unsolicited review of a book of your choice; or, if you would welcome being assigned a book to review, please write us, telling us your expertise and your predilections.