Development of a metamodeling tool for optimisation of control system

Abstract

This project focuses on the development of a metamodeling tool for use in the optimization of a control system. This integrated and user friendly metamodeling tool is to aid a designer in producing an accurate metamodel of the original system for optimization purposes. In metamodeling or model reduction, a simpler approximate of the actual simulation model of a system is determined which is called a metamodel. Metamodeling has been successfully applied in many fields where complex computer models of the actual system exist but running the simulation of the model costs a considerable amount of time. Metamodel is used in place of the actual simulation model to reduce the computational cost involved. The metamodeling system for control system optimization will be integrated into a software package with Graphical User Interface (GUI). The software used in developing this metamodeling tool is the MATLAB Graphical User Interface Development Environment (GUIDE) in MATLAB version 7.6. The type of metamodel used is the Radial Basis Function (RBF) Neural Network (NN).