Pakistan Research Repository

STATISTICAL CONTROL CHARTING ENHANCEMENTS UNDER DIVERSE CONDITIONS

Ali, Tasneem (2003) STATISTICAL CONTROL CHARTING ENHANCEMENTS UNDER DIVERSE CONDITIONS. PhD thesis, University of Karachi, Karachi.

[img]HTML
17Kb

Abstract

Statistical Process Control (SPC) is an effective tool of monitoring the process and detecting the causes, which shifts the process from the target. One of the SPC’s important tools is control chart. The process failure mechanism follows a weibull Distribution. The economic aspect introduced in the control chart makes it more efficient to monitor the performance of the system. Most quality control experimenters wish to study the quality characteristic having some symmetrical distributions due to readily available control charts. In reality, practitioners must focus on the true underlying distribution of the quality characteristic, which in most cases are asymmetristic, which in most cases are asymmetrical. The study is envisaged to monitor the process economically during on-line and while it is off-line The economic aspect of control chart depends on three design parameters; sample size, sampling interval and width of the control chart. We have applied pattern search technique to estimate the optimal economic design parameters. This technique helps us to choose the best economic design parameters, for which the cost function is minimum. It is also shown that choosing any possible combination of parameters for two-parameter weibull distribution but keeping the mean rate fixed, does not affect the optimization of the economic design parameters. The study is further extended to find the optimal economic design parameters in various situations of production being continued or ceased during the search of the assignable cause and the repair of the process The sensitivity analysis of the design parameters is carried out by changing the magnitude of shift, cost and time input variables. It is also based on types of errors, OC curves and the Average Run Lengths (ARLs) The second part of the study focuses on the offline quality control. In this part, we have enumerated the loss incurred to the producer as well as to the customer when the products are shipped to the customer with or without inspection. In this case the quality characteristic is a truncated weibull variable and the cost function depends only on one decision variable. The sensitivity analysis is carried out for the loss function on the shift. The comparative study is also based on Type I, Type-II errors, OC curves and ARLs

Item Type:Thesis (PhD)
Uncontrolled Keywords:statistical control, statistical process control (spc), average run lengths (arls), quality system, quality control, quality cost
Subjects:Physical Sciences (f) > Mathematics(f5) > Statistics(f5.2)
ID Code:1219
Deposited By:Mr. Muhammad Asif
Deposited On:04 Jan 2007
Last Modified:04 Oct 2007 21:05

Repository Staff Only: item control page