 |
 |
|
 |
 |
|
 |
 |
 |
In factorial design, only the linear effects of the quantitative factors are studied. Response surface methodology allows you to study the quadratic effects of the factors (i.e., effects that differ depending on the level of the factors), making it well-suited to predictive modeling and optimization. The following types of response surface method designs are available:
Central Composite: A two level full or fractional factorial design is embedded in the central composite design, and additional center and axial points are also used in order to estimate curvature. This is the most commonly used response surface methodology design.
Box-Behnken: Each factor must have three levels, with one level being the center point between the high and low levels. This design type is useful in cases where setting all factors at extreme values simultaneously is undesirable (e.g., if setting all factors at the high level carries a risk of equipment damage or otherwise violates constraints).
The ReliaWiki resource portal provides more information about response surface method designs at http://www.ReliaWiki.org/index.php/Response_Surface_Methods_for_Optimization.
The information presented in these topics is not intended to be an exhaustive discussion of the software. Rather, it offers a summary of each design type along with notes on any special considerations for creation/use of the design and information on the types of analysis information and plots available. It is intended to be used in conjunction with the documentation on design folios.
© 1992-2016. ReliaSoft Corporation. ALL RIGHTS RESERVED.
 |
E-mail Link |