DocumentCode :
274643
Title :
Fast integrated system optimization and parameter estimation algorithms
Author :
Brdys, M.A. ; Lok, K.
Author_Institution :
Birmingham Univ., UK
fYear :
1991
fDate :
25-28 Mar 1991
Firstpage :
632
Abstract :
It is proposed to use a gradient approximation in efficient second order nonlinear programming algorithms like conjugate gradient and variable metric techniques to obtain the second order ISOPE algorithms. Three alternative approaches based on this idea are investigated and verified through numerical simulation. Suitable line search algorithms are designed in order to appreciate the quality of the gradient approximation at a given iteration. Sufficient general purpose conditions for the algorithms convergence are derived and discussed
Keywords :
approximation theory; convergence of numerical methods; optimisation; parameter estimation; conjugate gradient; convergence; gradient approximation; integrated system optimization and parameter estimation algorithms; iteration; line search algorithms; second order ISOPE algorithms; second order nonlinear programming algorithms; variable metric techniques;
fLanguage :
English
Publisher :
iet
Conference_Titel :
Control 1991. Control '91., International Conference on
Conference_Location :
Edinburgh
Print_ISBN :
0-85296-509-5
Type :
conf
Filename :
98523
Link To Document :
بازگشت