Multiobjective Design Optimization Using Adaptive Classification

Abstract

Definitions of a design space and an objective space for performing multi-purpose design optimization of a product are received in a computer system on which the design optimization application module is installed. The design space is defined by a plurality of design variables, and the purpose space is defined by a plurality of design objectives. A first set of design alternatives is initially selected within the design space. The first set of design alternatives is selected in the design space. Each of the first group was evaluated in the purpose space for the non-dominant. The design space is partitioned into first and second regions using a multidimensional space partitioning scheme (eg, SVM). The first region is the portion of the design space that contains all non-dominant design alternatives, while the second region contains the remainder of the design space. Select the second set of designs in the first area. Each of the second group and the existing non-dominant design candidates were evaluated for non-dominance. Multipurpose design optimization repeats the segmentation and evaluation until an end condition is reached.

Description

technical field [0001] The present invention relates generally to engineering design optimization, and more particularly to multi-objective design optimization using adaptive classification to select design alternatives. Background technique [0002] Today, computer-aided engineering (CAE) has been used to support engineers in tasks (eg, analysis, simulation, design, manufacturing, etc.). In routine engineering design programs, CAE analysis (e.g., finite element analysis (FEA), finite difference analysis, meshless analysis, computational fluid dynamics (CFD) analysis, modeling for reducing noise vibration harshness (NVH) state analysis, etc.) have been used to evaluate the response (eg, stress, displacement, etc.). Using the car design as an example, a particular make or design of a car is analyzed using FEA to obtain the response due to certain loading conditions. Engineers will then try to improve the car design by modifying certain parameters or design variables (for ex...

AI Technical Summary

Problems solved by technology

[0005] One problem with obtaining POF in multipurpose design optimization is the requirement to have a large number of experiments (i.e. unique design alternatives in the design space), which can be very expensive in terms of time and / or computational cost

Images