MULTI-OBJECTIVE MDO SOLUTION STRATEGY FOR MULTIDISCIPLINARY DESIGN USING modeFRONTIER

Design of complex engineering systems often involves multiple interacting disciplines and analyses. With the increase in computational power and sophisticated modeling techniques, the exploitation of the synergistic relationship that exists between coupled disciplines is becoming increasingly popular. While the interacting disciplines are trying to achieve one optimum system, the problem definition also often involves multiple conflicting objectives. The goal of this paper is to evaluate the application of a heuristic optimization strategy to solve tightly coupled Multidisciplinary Optimization (MDO) problems that involve multiple objectives. The multi-objective MDO problem is solved using the response surface methods and evolutionary algorithms available in modeFRONTIER [1]. A non-hierarchic tightly coupled MDO problem is formulated using Multidisciplinary Feasible (MDF) [2,3] formulation and solved using Non-dominated Sorting Genetic Algorithm (NSGA-II) [4]. NSGA-II allows the generation of evenly distributed Pareto designs in a faster and more efficient manner. The iterative convergence of coupling variables called System Analysis (SA) is reduced by utilizing response surface approximations, with subsequent update and refinement, thereby greatly increasing computational efficiency. The solution strategy is applied to a mathematical MDO test problem and the results are discussed