Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.742289
Title: Information theory in quality engineering
Author: Svensson, Bolennarth
Awarding Body: University of Strathclyde
Current Institution: University of Strathclyde
Date of Award: 1998
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
This thesis presents the results of research into a universal theory for quality techniques. The unique contribution that is made is twofold: • A new quality metric is proposed; • An integrating perspective to quality engineering is introduced through the application of information theory. The quality metric is designed as an information distance, measuring the difference between two probability density functions. The two distributions are the actual outcome of a running process and the expected outcome, i.e. the target distribution. The target distribution makes it possible to integrate the quality losses into the metric. The metric may be adapted to the state of knowledge of the process studied. The new quality metric is applicable to any process, be it a product processor an administration process. The information distance metric makes the analysis procedures uniform for all types of quality characteristics. A function based process documentation makes information theory generally applicable to quality engineering. The function description makes it possible to visualize poor quality as a surplus of information. All quality techniques aim at minimizing the information content in the system. Quality engineering in general may be expressed as an activity to stop surplus information flow reaching the process result. There is a natural focus on noise, i.e. influencing factors that are out of control of the user, affecting the systems. This focus is in robust design developed through a process performance perspective rather than an experimental design perspective. An effort addressing a product process subject to improvement has to be discriminated from an effort addressing the efficiency of the experimentation process used to study the product process. The present work is of pioneering character. Thus it opens a new area of research. Areas of further research are indicated.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.742289  DOI:
Share: