توظيف نهج سطوح الاستجابات المتعددة المرتبطة في تصميم المعلمة الحصين مع تطبيقات عملية
Author name:
فرح عصام حسن
Supervisor name:
ابتسام مصطفى كمال
General topic:
Administration and Economics
Specific topic:
Statistics
Degree:
Doctorate
University:
University of Baghdad - Faculty Of Administration And Economics - Department Of Statistics
Language:
Arabic
University location:
Baghdad
First pages:
07T3948 - p.pdf
Abstract:
هذه الاطروحة كانت محاولة لايجاد اساليب كفوءة يتم فيها معالجة حالة الاستجابات المتعددة المرتبطة في تصميم المعلمة الحصين، بتعميم نهج سطوح الاستجابة وتوظيفه في خمسة اساليب مقترحة اخرى، في كل منها تم الاعتماد على دالة هدف مختلفة تعالج حالة معينة، والدوال المو | The purpose of this study was to find efficient techniques that extend the application of response surface methodology to the multiple robust parameter design, for cases in which multiple responses are correlated. The study recommends the use of five techniques, in each one a different function was employed. These functions are : 1. Squared loss function2. Principal component function3. Canonical correlation function4. Multiple correlation function5. Partial correlation function In the five suggested techniques the model of Quesada & Casteillo which is a generalization of Vining and Myres model for the one response case, was considered. The model was assumed to follow the assumptions of the seemingly unrelated regression equations system. And Zenller’s two stage technique was employed to estimate its parameters. The five suggested techniques were employed in five different experiments for different fields. Then the results were compared with the results of Quesada & Casteillo approach for multiple robust parameter design. The comparison was based on criteria of the average mean square error, which was the lowest at the most optimal solutions produced from the suggested techniques and for the following applications : 1. Optimization and robustness of the High - performance Liquid Chromatography. 2. Optimization and robustness of the carry and release function of a coated drug.3. Optimization and robustness of the computers network efficiency.4. Improvement and robustness of the quality of coffee treated by a new technique.5. Optimization and robustness of the function of a printing machine in adding colored inks.
