摘要:

A preliminary control chart based on individual observations for detecting a shift in either the mean or variance or both is developed from a likelihood ratio test. The advantages over use of theXcontrol chart and moving range control chart are improved detection of sustained shifts, in addition to improved diagnostic information indicating multiple shifts and whether an out-of-control signal is primarily due to a shift in the mean or variance or a combination. Approximate control limits, determined from simulation, are given for the proposed control chart.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979677

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

Traditional methods of statistical process control (SPC) assume that observations of a quality characteristic are uncorrelated. This paper examines the case in which unacceptable variability in a product at one station in a production sequence will cause unacceptable variability in a product at the succeeding station. Thisserial dependencemeans that conventional SPC may be ineffective. For example, serially-dependent processes with attribute data may be nonstationary, demonstrating a geometric increase in the probability of occurence of a defect. This increase can be arrested by placing control points in the chain, with respect to the variability of the process at each work station. A control point is an inspection or correction station or both, that will re-establish the product attribute to a desirable level at that point in the sequence. Some of the resulting segments will contain stations with such low probability of defect that a correlated effect over an abbreviated segment may behave according to distributions that can be controlled by conventional SPC methods. Other segments may magnify short trends to create out-of-control conditions. These segments can be controlled by regression control. This paper develops a control system that maintains stability of serially-dependent processes irrespective of the number of stations or basic defect rates of the individual work stations in the sequence.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979678

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

The problem of monitoring autocorrelated process data is considered. A Markov chain approach is used to determine the performance of several forecast-based monitoring schemes for various types of process shifts and evaluation criteria. The performance of the monitoring schemes is shown to depend on the process model, type of shift, and evaluation criteria. The combined exponentially weighted moving average-Shewhart (CES) control chart applied to appropriate one-step-ahead forecast errors is recommended for general use, and a design procedure is provided.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979679

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

Autocorrelation in process measurements results in false alarms when traditional control charts, such asX̄andRcharts, are applied in process monitoring. A popular approach proposed in the literature is to use time series models to remove the autocorrelation, and then to apply control charts to the residuals, or prediction errors. However, when a time series model is used to remove the autocorrelation, the change patterns may be severely altered, making the interpretation of a process change and root cause analysis erroneous. This paper develops charts for the comparison of change patterns for two types of additive process faults before and after applying time series modeling. The process changes are sustained mean shifts and sporadic spikes. These charts provide improved understanding of the characteristics of autocorrelated processes.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979680

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

A projection approach to designing multivariate control schemes is used that simplifies the construction and understanding of these charts. The method is applied to construct charts that are sensitive to specific assignable causes such as a mean shift that affects only a subset of the variables and a model-fixed assignable cause that can occur when one variable causes another. The projection results in a chart that is easy to compute and design and it has greater sensitivity than a chi-square chart for the assignable cause under study. Furthermore, the chart can be analyzed using the convenient central and noncentral chi square distributions when the process is on-target and off-target, respectively.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979681

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

Variables control charts implement process state testing by separately assessing the measures of central tendency and spread. This work, based on the ideas of bivariate process control, analyzes whether the standard charting method could be considered as optimal and presents an alternative mode of process evaluation—biparameter description in terms of the joint distribution of sample averages and standard deviations. Thus it becomes possible to benefit from the advantages of this approach and at the same time to overcome its main shortcoming, namely lack of diagnostic ability. The proposed singleX̄chart with adaptive control limits provides a reliable tool for recognizing process state and indicating the parameter responsible for an out-of-control state. Additionally, the proposed method does not require a constant sample size during data collection.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979682

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

Recently several authors have discussed how one can solve the three dual response problems proposed by Taguchi (target is best, larger is better, and smaller is better) without resorting to signal-to-noise ratios. The most recent articles have all relied on nonlinear programming techniques to obtain the actual solution. While these techniques certainly work, it is also possible to solve the dual response problems using the more familiar technique of direct function minimization. We demonstrate this using the Nelder-Mead simplex procedure. We also propose slightly different formulations of the problems, which seem to be more realistic. Examples for each type of dual response problem are presented.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979683

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

Desirability functions have been used extensively to simultaneously optimize several responses. Since the original formulation of these functions contains non-differentiable points, only search methods can be used to optimize the overall desirability response. Furthermore, all responses are treated as equally important. We present modified desirability functions that are everywhere differentiable so that more efficient gradient-based optimization methods can be used instead. The proposed functions have the extra flexibility of allowing the analyst to assign different priorities among the responses. The methodology is applied to a wire bonding process that occurs in semiconductor manufacturing, an industrial process where multiple responses are common.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979684

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

In the continuous process industries, multiple grades of the same product are often made on the same production equipment. Key product properties for each grade are usually monitored separately because of inherent differences among grades. Although highly effective in detecting changes that affect individual grades such as changes in grade-specific raw material properties, this approach is usually slow in detecting changes common to all grades such as a drift caused by production equipment or changes in common raw materials. The case study of this paper illustrates a simple method of combining and monitoring data from multiple grades allowing speedy detection of changes affecting all grades and yet still permitting differentiation of overall process shifts from grade-specific shifts.

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979685

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor

摘要:

A test for normality is available that makes use of the fact that only for the normal distribution are the averagex̄and the sample standard deviation s independent. Users are cautioned that the test is not sensitive to nonnormal distributions that are symmetrical. Normal probability paper is explored as a tool for use, in addition to the test, when specifically testing for normality

ISSN:0022-4065    

DOI:10.1080/00224065.1996.11979686

出版商:Taylor&Francis    

年代:1996

数据来源: Taylor