A New Robust Estimation Approach to Dual Response Surface Models

Robust parameter design is one of the methods used to determine the optimum operating conditions of systems. In robust parameter design, dual response surface methodology plays a crucial role in identifying optimal settings that minimize variability while maintaining desired performance. The methods generally used in dual response surface models are based on the assumption of normal distribution. Nonetheless, using techniques that rely on traditional normality assumptions to model non-normally distributed data often leads to the failure of quality improvement processes. In cases where there is deviation from normality, using robust estimators in response surface models for mean and variance has become very popular in recent years. This study focuses on skewed data, and thanks to the proposed approach, it is aimed to produce more robust results than the models in the literature by using the BS82 robust M estimator instead of standard deviation, with modelling for confidence intervals. The main advantage of the proposed approach is that it provides a robust solution that considers the skewed nature of the experimental data distribution. To demonstrate the validity of the approach, the newly developed model has been examined printing process data is frequently discussed in the literature.