摘要:

In order to provide guidelines for inspection planning that minimize the total cost of inspection, Dr. Deming developed thek1,k2inspection models. To date, theoretical expected values for the total cost have been published for 100 percent inspection, zero inspection and acceptance sampling, but theoretical variances and probability distributions have not. In this paper, the theory of compound distributions is used to develop the probability distribution of total cost for 100 percent inspection and zero inspection and to obtain the expected value and variance of total cost for 100 percent inspection, zero inspection, and acceptance sampling. The models are then verified with the use of computer simulation.

ISSN:0022-4065    

DOI:10.1080/00224065.1995.11979617

出版商:Taylor&Francis    

年代:1995

数据来源: Taylor

摘要:

While statistically designed experiments have been employed extensively to improve product or process quality, they have been used infrequently for improving reliability. In this paper, we present a case study which used an experiment to improve the reliability (or lifetime) of fluorescent lamps. The effect of three factors from among many potentially important manufacturing factors was investigated using a fractional factorial design. For fluorescent lamps, failures occur when their luminosity or light intensity falls below a certain level. An interesting feature of this experiment is the periodic monitoring of the luminosity. The paper demonstrates how the luminosity's degradation over time provides a practical way to improve fluorescent lamps which are already highly reliable. Recommendations based on the experiment's results suggest that nearly a 70% improvement can be realized.

ISSN:0022-4065    

DOI:10.1080/00224065.1995.11979618

出版商:Taylor&Francis    

年代:1995

数据来源: Taylor

摘要:

A signal from a multivariate control chart may be difficult to explain. One of the techniques used to facilitate this action is to chart the principal components, but if the components are not easily interpreted the problem remains. This paper expands upon previous work and provides an interesting bivariate setting in which the principal components have meaningful interpretations. When monitoring a process with paired measurements on a single sample, the principal components of the corresponding correlation matrix actually represent the characteristics of interest for process control. Further, the correlation coefficient between the original variables is the only additional information needed to describe the condition of the process.

ISSN:0022-4065    

DOI:10.1080/00224065.1995.11979619

出版商:Taylor&Francis    

年代:1995

数据来源: Taylor

摘要:

Schneider, Kasperski, and Weissfeld (1993) proposed a censored sampling approach for analyzing unreplicated factorial and fractional factorial experiments. This paper presents a FORTRAN program that implements the Schneider, Kasperski, and Weissfeld methodology to determine thetstatistics used for testing the significance of the effect estimates (contrasts), and then performs the test for all of the effects.

ISSN:0022-4065    

DOI:10.1080/00224065.1995.11979620

出版商:Taylor&Francis    

年代:1995

数据来源: Taylor

ISSN:0022-4065    

DOI:10.1080/00224065.1995.11979621

出版商:Taylor&Francis    

年代:1995

数据来源: Taylor

ISSN:0022-4065    

DOI:10.1080/00224065.1995.11979622

出版商:Taylor&Francis    

年代:1995

数据来源: Taylor

ISSN:0022-4065    

DOI:10.1080/00224065.1995.11979623

出版商:Taylor&Francis    

年代:1995

数据来源: Taylor

ISSN:0022-4065    

DOI:10.1080/00224065.1995.11979624

出版商:Taylor&Francis    

年代:1995

数据来源: Taylor

ISSN:0022-4065    

DOI:10.1080/00224065.1995.11979626

出版商:Taylor&Francis    

年代:1995

数据来源: Taylor

ISSN:0022-4065    

DOI:10.1080/00224065.1995.11979627

出版商:Taylor&Francis    

年代:1995

数据来源: Taylor