Title :
PSO Based Surrogate Model Steady State Optimization with Its application
Author :
Li, Xiugai ; Huang, Dexian
Author_Institution :
Dept. of Autom., Tsinghua Univ., Beijing
Abstract :
RBF neural network based surrogate model is constructed from process simulator and particle swarm optimization (PSO) strategy is discussed to solve this nonlinear programming problem with some output variables are unmeasured. Performance of maximum yield rate for a hydrolysis of the propylene oxide reaction is developed and the constrained PSO can give the best value of the steady state with different inflow rate and temperature. Comparison results with the evolutionary strategy (ES) show the efficiency of this new intelligent optimization method, it provides a practically method for industry process optimization
Keywords :
nonlinear programming; particle swarm optimisation; radial basis function networks; PSO based surrogate model; RBF neural network; constrained PSO; evolutionary strategy; hydrolysis; intelligent optimization method; nonlinear programming problem; particle swarm optimization; propylene oxide reaction; radial basis function; steady state optimization; Automatic programming; Automation; Electronic mail; Intelligent control; Minimax techniques; Neural networks; Optimization methods; Particle swarm optimization; Steady-state; Temperature; Evolutionary Strategy; Particle Swarm Optimization; Radial Basis Function; Surrogate Model;
Conference_Titel :
Intelligent Control and Automation, 2006. WCICA 2006. The Sixth World Congress on
Conference_Location :
Dalian
Print_ISBN :
1-4244-0332-4
DOI :
10.1109/WCICA.2006.1714354
