Resumo

SALMORAN-LOPEZ, Edgar; RIOS-LIRA, Armando Javier  e  PANTOJA-PACHECO, Yaquelin Verenice. Computational program to generate alias structures of mixed level fractional factorial designs. Ing. invest. y tecnol. [online]. 2020, vol.21, n.4, 00003.  Epub 20-Nov-2020. ISSN 2594-0732.  https://doi.org/10.22201/fi.25940732e.2020.21.4.031.

The purpose of this document is to show the development of a program encoded in MATLAB that aims to facilitate the construction of alias structures of mixed levels fractional designs. The program can construct alias structures of mixed levels fractional designs of 2 to 9 factors without restricting the number of runs. In addition, it performs a search for strong correlations between main effects, which prevents the experimenter from selecting a design in which the main effects are confused with each other. Imagine that the experimenter needs to build the alias structure of a 9 factor design matrix with 30 runs, which involves basic math operations, matrix operations and the analysis of a correlation matrix of size 129 × 129, this would take at least 3 hours if he master the method and also use a spreadsheet. Therefore, this program aims to facilitate the construction of alias structures and cancel the uncertainty of calculation and interpretive errors, in addition to the reduction of time. The experimenter will be able to know the alias structure of a fraction of a design of mixed levels in a simple way, being able to concentrate on the one that provides the most information of the terms of interest. This work is limited to showing the algorithm that was codified in MATLAB for the construction of alias structures of mixed levels fractional factorial designs, but not the efficiency of the method to show these structures.

Palavras-chave : Designs; mixed; structures; alias; program.