2014
EMSS 2014 Proceeding
A multimodal optimization method for simulation systems
Authors: Zhizhao Liu, Wei Li, Ming Yang
Abstract
Considering multimodal optimization problem and potentially run times for continuous simulation systems, Particle Swarm Optimization (PSO) method based on Support Vector Machine (SVM) and Cluster Analysis (CA) is proposed. SVM, a global metamodel of simulation system, is built and PSO algorithm is used to search for local optimal points based on SVM. To determine local optimums, the population individuals of PSO are classified by CA. Consequently, six typical multimodal function optimization problems are selected to verify the optimization performance.
I3M Scientific Sponsors
I3M Industrial Sponsors
I3M Media Sponsors
