Performance Analysis of Nelder - Mead and A Hybrid Simulated Annealing for Multiple Response Quality Characteristic Optimization

In many real world industrial situations, quality of a product is considered to be a composite of a family of properties or multiple responses, which can often be interacting or correlated with one another, and nearly always measured in a variety of units. Optimal combination of properties rarely results from individual optimal condition of each response, and there is always a trade off between these responses. Optimization of multiple responses is always a challenge for researchers and practitioners. In this research paper, comparative study of the performance of two optimization techniques, Nelder-Mead simplex (NM) and hybrid Nelder-Mead simplex with simulated annealing (SIMSA) is discussed. The results indicate superiority of SIMSA as compared to NM in varied case situations.