Title :
A constrained simultaneous perturbation stochastic approximation algorithm based on penalty functions
Author :
Ang, I-jengw ; Spall, Jamecs
Author_Institution :
Appl. Phys. Lab., Johns Hopkins Univ., MD, USA
Abstract :
We present a stochastic approximation algorithm based on the penalty function method and a simultaneous perturbation gradient estimate for solving stochastic optimization problems with general inequality constraints. We also presents a very general convergence result for the proposed algorithm
Keywords :
approximation theory; constraint theory; convergence of numerical methods; function approximation; optimisation; constrained optimization; convergence; function approximation; gradient estimation; inequality constraints; penalty function; perturbation; stochastic approximation; Algorithm design and analysis; Approximation algorithms; Concrete; Constraint optimization; Convergence; Cost function; Laboratories; Optimization methods; Physics; Stochastic processes;
Conference_Titel :
Intelligent Control (ISIC), 1998. Held jointly with IEEE International Symposium on Computational Intelligence in Robotics and Automation (CIRA), Intelligent Systems and Semiotics (ISAS), Proceedings
Conference_Location :
Gaithersburg, MD
Print_ISBN :
0-7803-4423-5
DOI :
10.1109/ISIC.1998.713704
