Overview
Français
ABSTRACT
The Taguchi method has enriched experimental design methods by considerably improving the full and fractional factorial designs. It has simplified the experimental protocol in order to highlight the factor effects on the response. The Taguchi method stands out by the considerable reduction of the number of tests whilst maintaining a high precision level. It defines the model as a key element of the experimental design strategy. The experimenter freely chooses the factors and interactions to be studied according to the model they offer in close adequation with their objectives.
Read this article from a comprehensive knowledge base, updated and supplemented with articles reviewed by scientific committees.
Read the articleAUTHOR
-
Rachid SABRE: MAIS Laboratory - Établissement national d'enseignement supérieur agronomique de Dijon (ENESAD)
INTRODUCTION
Taguchi's method enriches design of experiments methods by bringing a considerable improvement to the complete and fractional factorial designs detailed in the dossier. .
Its aim is to simplify the experimental protocol in order to highlight the effects of factors on the response, which may be a variable in an agri-food process or even the measurement of product quality.
Taguchi's method is characterized by a significant reduction in the number of trials, while maintaining good precision. It places the model as a key element of the experimental design strategy. The experimenter is free to choose the factors and interactions to be studied according to the model he or she proposes, in close alignment with his or her objectives.
The place of each factor in the Taguchi design is important, and is chosen according to the difficulty of realizing the factor in the experiment. This place in the plan will enable the factor that is the most difficult to realize to make the fewest possible level changes. In this way, factors can be grouped by degree of difficulty of realization.
Taguchi's method initially met with success in the industrial sectors and, in particular, in the agri-food sector, and then aroused the interest of the statistical community for wider development and study.
In this file, we begin by defining orthogonal tables with respect to a model. We also give some properties concerning these tables, which we will later use to choose the Taguchi table that best suits the problem under study. We then study Taguchi's method, showing how to choose the appropriate orthogonal table from among the existing ones, and how to calculate the model coefficients needed to highlight the effects of the factors under study.
The case where factors form 2 groups of different natures (internal factors and external factors) is addressed by studying the product design crossing two experimental designs and calculating the coefficients of the product model.
Some Taguchi tables are given as examples in the Bibliography and Appendices tab.
The Ultimate Scientific and Technical Reference
This article is included in
Food industry
This offer includes:
Knowledge Base
Updated and enriched with articles validated by our scientific committees
Services
A set of exclusive tools to complement the resources
Practical Path
Operational and didactic, to guarantee the acquisition of transversal skills
Doc & Quiz
Interactive articles with quizzes, for constructive reading
Design of experiments
Orthogonal tables with respect to a model
Bibliography
General books and articles
Calculation software
(Non-exhaustive list)
Nemrod, software built at LPRAI, University of Aix-Marseille.
Sas, publisher Institute.
...
The Ultimate Scientific and Technical Reference
