A Comparative Study for Approximation Techniques to Nonlinear Optimization Problems

Abstract

Constrained nonlinear programming problems often arise in many engineering applications and much of today's engineering analysis consists of running complex computer codes. Despite a steady increase in computing power, the complexity of engineering analyses seems to advance at the same rate. The use of statistical techniques to build approximations of expensive computer analysis codes pervades much of today's engineering design. These statistical approximations, or meta-models, are used to replace the actual expensive computer analyses, facilitating multidisciplinary and concept exploration.