METHOD AND SYSTEM FOR MULTI-OBJECTIVE EVOLUTIONARY ALGORITHM BASED ENGINEERING DESIGN OPTIMIZATION

摘要

PPROBLEM TO BE SOLVED: To provide systems and methods of conducting multi-objective evolutionary algorithm (MOEA) based engineering design optimization of a product (e.g., automobile, cellular phone, etc.). PSOLUTION: Particularly, there is provided an archive configured for monitoring the progress and characterizing the performance of the MOEA based optimization. Further, an optimization evaluation indicator is created using the archive's update history. The optimization evaluation indicator is used as an evaluation of the current state of the optimization. Finally, a stopping or termination criterion for the MOEA based optimization is determined using an evaluation derived from the optimization performance indicators. For example, a confirmation of a "knee" formation has developed in the optimization evaluation indicators. The optimization evaluation indicators include, but are not limited to, consolidation ratio, improvement ratio, hypervolume. PCOPYRIGHT: (C)2011,JPO&INPIT