摘要:

This article relates lessons that are useful in the analysis of uncertainties in a complex measurement process, through a case study. It distinguishes between the settings for confidence intervals and those for prediction intervals. It illustrates testing of an assumed variance, shows how to separate the variance contributors in a calibration curve, discusses the use of independence of variance components, gives first-order approximations for combining random errors, and emphasizes the necessity of writing all the contributors to uncertainty in a single mathematical expression.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979632

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

Traditional applications of control charts use fixed sampling interval (FSI) charts in which the time interval between samples is fixed. Recently, more efficient variable sampling interval (VSI) charts have been developed in which the next observation is taken sooner than usual if there is an indication that the process is operating off the target value. It has been found that VSI charts can detect process shifts faster than FSI charts. The underlying assumption behind most statistical process control charts is that the observations from the process are independent. However, there are many practical situations in which the observations cannot be treated as independent, particularly if they are closely spaced in time as may be the case with a VSI chart. In this paper successive values of the process mean are modeled using a first-order autoregressive time series model, and observations are modeled as the mean plus a random error. A Markov process model is used to study the properties of the FSI and VSIX̄charts under this model. Numerical results show that correlation between successive means has a significant effect on the properties of both FSI and VSIX̄charts. When correlation is present, the VSIX̄chart will detect process shifts faster than the FSIX̄chart, but for high correlation there is little difference between the performance of the two charts.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979633

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

We address a problem of increasing practical interest: control charting in the absence of an assumed normal distribution. We analyze the sampling uncertainties inherent in selecting control limits from the empirical distribution of a large sample of whatever statistic is to be charted. We then relate the selections of the number of preliminary observations and control limits to on-target and off-target performance measures. The methodology is nonparametric, requiring only that the plotted statistics be independent and come from the same distribution.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979634

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

The productivity of an industrial processing unit often depends on equipment that changes over time. These changes may not be consistent, and, in many cases, may appear to occur in stages. Although changes in the process levels within each stage may appear insignificant, they can be substantial when monitored across the various stages. Standard process control procedures do not perform well in the presence of these step-like changes, especially when the observations from stage to stage are correlated. This paper presents an alternative multivariate control procedure for monitoring a process under these conditions. The methodology is based on a double decomposition of Hotelling'sT2statistic.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979635

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

This case study, involving a continuous processing operation of a plastic material, provides a simple, intuitively attractive evaluation of the potential reduction in variability that can be achieved by blending materials after production. The evaluation is based exclusively upon past data onunblendedmaterials. The evaluations were complicated by the fact that measured variability was inflated by analytic error. The paper isnotan endorsement of blending as a strategy for improving consistency. In fact, this evaluation showed that the blending that was practical in this application would provide only minimal improvements; thus the idea of blending was abandoned.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979636

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

This paper presents a new methodology for analyzing multiresponse experiments. The methodology consists of computing confidence regions for the stationary points of quadratic responses and confidence cones for the direction of maximum improvement for linear responses. The stationary points are constrained to lie within the region of experimentation. It is shown that the confidence regions depend on the value of the Lagrange multiplier of the region's constraint. The value of the Lagrange multiplier is found by solving the Karush-Kuhn-Tucker optimality conditions. Then, nonlinear optimization problems are set up and solved for obtaining experimental points that lie inside all the confidence regions, cones and constraints. Robust process design examples illustrate the methods proposed. The examples address the“target is best”and“larger the better”cases.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979637

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

Ammunition deterioration during storage has considerable economic consequences. Thus, a reliable prediction model for the ammunition deterioration rate is necessary for long-term procurement and maintenance planning. The purpose of this paper is to formulate a prediction model for ammunition deterioration rates in terms of concurrent characteristics such as depot condition and vendor information using a random effects growth curve analysis. The resulting prediction model can be used to determine the appropriate time for reorder or renovation of ammunition before the quality reaches unacceptable levels. A two-stage analysis is used to estimate parameters involved in the prediction model. Necessary estimation methods are discussed and an example is given to illustrate the procedure.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979638

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

The modulated power law process (MPLP) is a three-parameter stochastic point process model that can be used to describe the failure times of a repairable system. While the power law process implies that a system is in exactly the same condition just after a repair as it was just before the failure, the MPLP allows for the system to be affected by the failure and repair. We describe an algorithm for obtaining the maximum likelihood estimates (MLEs) of the three model parameters. Asymptotic results are used to give approximate confidence intervals and hypothesis tests for the parameters. A simulation study indicates that the confidence intervals and hypothesis tests have approximately the nominal level.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979639

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

Process capability analysis has focused primarily on univariate capability indices. We take a broader view and introduce the concept ofexploratorycapability analysis, an essentially multivariate activity aimed at capability improvement rather than assessment. We present three illustrative examples involving multiple dimensional measurements where the within-part inspection plans are structured like lattices in rectangular, cylindrical, or other geometries.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979640

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

Presented is a charting technique for displaying process changes when several process measurements are important. The technique can be used for nominal, ordinal or variable data. Process performance on all important measurements is presented on a single chart. The chart is useful for management to link poor performance in outcome variables to measures taken within the process. This helps management isolate causes of poor performance on outcome variables.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979641

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor