Stanford University, Palo Alto, California, 94305

摘要

A new multi-objective optimization algorithm called Pareto Simplex Search (PSS) is proposed. The new scheme is a (u/p+X) Evolution Strategy which is based on the application of a random-step reflection operator on a randomly selected simplex among a population of individuals. PSS has been applied to the problem of aerodynamic shape optimization of unguided projectiles with a stability criteria and the aerodynamic efficiency as the two objectives. The problem has also been solved by a multi-objective real-coded genetic algorithm (NSGA-II) and the results are compared.