Penalty Function Optimization in Dual Response Surfaces Based on Decision Maker’s Preference and Its Application to Real Data

The dual response surface methodology is a widely used technique in industrial engineering for simultaneously optimizing both the process mean and standard deviation functions of variables. Many optimization techniques have been proposed to optimize two fitted that include penalty function method (PM). PM has shown be more efficient than some existing methods. However, drawback it does not specific rule determining constant; thus, practice, practitioners will find this difficult since depends on subjective judgments. Moreover, most methods, sample often use non-outlier-resistant estimators. ordinary least squares (OLS) also usually estimate parameters functions. Nevertheless, many statistics are aware OLS procedure classical easily influenced by presence outliers. Alternatively, instead using those we propose high breakdown highly robust MM-mean, MM-standard deviation, MM regression estimators overcome these shortcomings. We new incorporates systematic determine constant. call based decision maker’s (DM) preference structure obtaining constant, denoted as PMDM. performance our investigated Monte Carlo simulation study real examples employ symmetrical factorial design experiments (DOE). results signify PMDM compared other commonly methods study.