Frontiers in Statistical Quality Control 5

ISBN-13: 9783790809848 / Angielski / Miękka / 1997 / 297 str.

Hans-Joachim Lenz;Peter-Theodor Wilrich

Opis

A collection of contributions on statistical quality control, stressing methodology and successful industrial applications. The papers are classified under four main headings: sampling inspection, process quality control, data analysis and process capability studies, and experimental design.

Spis treści

1: Sampling Inspection.- An Adaptive Sampling System with a Fuzzy Controller.- Determination of Sample Sizes for Double Sampling Attribute Plans.- Unbiased Estimation of Generalized Moments of Process Curves.- On the Non-Robustness of Maximum-Likelihood Sampling Plans by Variables.- Normal Approximations to the Distribution Function of the Symmetric Beta Distribution.- 2: Process Quality Control.- A Mathematical Framework for Statistical Process Control.- A Review of Statistical and Fuzzy Quality Control Charts Based on Categorical Data.- Efficient Estimation of Control Chart Parameters.- Control Charts for Dependent and Multivariate Measurements.- On EWMA Charts for Time Series.- Statistical Process Control for Autocorrelated Processes: A Case-Study.- Quality Control of a Continuously Monitored Production Process.- Group Sequential Design with Delayed Observations for Selecting One of Two Processes in a Production System.- Design of the ($$ \overline x $$, s) Control Chart Based on Kullback-Leibler Information.- 3: Data Analysis And Process Capability Studies.- Measurement Error Effects on the Performance of Process Capability Indices.- A New Approach for Describing and Controlling Process Capability.- On the Use of Field Failure Data for Repairable Systems to Identify Sources of Variation.- 4: Experimental Design.- Control Chart Method for Analyzing Staggered Nested Data.- Prediction Properties of the Process Variance Using the Combined Array.- Joint Analysis of Mean and Variance Function Based on Second Order Polynomials.- Dependability Improvement Through Unreplicated Experimental Designs.