Paper Authors

Abstract

NOTE: The first page of text has been automatically extracted and included below in lieu of an abstract

Session 2632

A Novel Graphical Technique for Selection of a Robust Design Point

José G. Colom*, R.R. Barton**, Lynn Carpenter**, Rafael Rodríguez Solís*

*University of Puerto Rico at Mayaguez/ **Penn State University

I. INTRODUCTION

The concept of robust design was introduced by Genichi Taguchi, and popularized in the U.S. following his visit to Bell Laboratories in the mid-1980’s. Since this technique has become important in industry, the topic has been included in many undergraduate engineering statistics curricula. Unfortunately, the experiment design and data analysis issues are complex, and particularly difficult to teach at an undergraduate level. In this paper, a novel graphical technique to select a robust design point is presented. The technique can be used in many design settings, but here it is focused on microwave circuit design. The principle is very simple and easy to understand by students at an undergraduate level, and at the same time it is very powerful to quickly achieve an optimum design for a circuit or system. The authors have implemented the technique for several microwave passive structures, but any optimum circuit or system can be designed in the same manner. The technique can be introduced in any design course and one to two lectures are enough to present it without the need of an extensive statistical background.

Passive microwave structures are usually fabricated using microstrip technology. A microstrip circuit basically consists of a dielectric material sandwiched between two metal layers. The metal layer at the bottom is used as the ground plane, while the top metal layer is etched to a specific geometry that will provide the desired frequency response (i.e. filter, directional coupler, power splitter, taper line, etc). The input variables for the microwave passive structures are usually the physical dimensions, while the output variables are frequency dependent parameters such as1 VSWR, gain, insertion loss, coupling, etc. In order to consider variations of physical parameters (i.e., spacing, lengths, widths, etc.) and their effect on the output response (VSWR, isolation, coupling, insertion loss, etc.) of a circuit, the analysis has to be performed in an organized manner. The integration of statistical tools in the area of microwave design or any other area of scientific investigation plays an important roll in the proper analysis of the problem under scrutiny. Different statistical tools, if properly used, can provide the designer with vital information with the minimum amount of resources. These runs, also known as experiments, are performed in a methodical and organized manner. The goal is to maximize the information that can be extracted from the original model with the minimum number of experiment. One statistical technique, created with that purpose in mind, is the Design of Experiments (DoE). With this technique, the input parameters of the system, or factors, are deliberately and systematically changed to capture variations in the output response. Input parameter settings are recorded along with the output response and the data is used to develop a metamodel or a "model of the model".

Proceedings of the 2001 American Society for Engineering Education Annual Conference & Exposition Copyright © 2001, American Society for Engineering Education

• Citation

• Format
  • APA
  • APA - LaTeX bibitem
  • MLA
  • MLA - LaTeX bibitem
  • Bibtex
  • EndNote - RIS

Rodríguez-Solís, R., & Carpenter, L., & Colom-Ustáriz, J., & Barton, R. (2001, June), A Novel Graphical Technique For Selection Of A Robust Design Point Paper presented at 2001 Annual Conference, Albuquerque, New Mexico. 10.18260/1-2--9615