Application of RBF Neural Network to Simplify the Potential Based Optimization

Authors:
Kanjian Zhang;Chunbo Feng
Affiliations:
Research Institute of Automation, Southeast University, Nanjing 210096, P.R. China;Research Institute of Automation, Southeast University, Nanjing 210096, P.R. China
Venue:
ISNN '07 Proceedings of the 4th international symposium on Neural Networks: Advances in Neural Networks, Part III
Year:
2007

Citing 5
Cited 0

Numerical methods for stochastic control problems in continuous time

Numerical methods for stochastic control problems in continuous time
Stability of an Adaptive Regulator for Partially Known Nonlinear Stochastic Systems

SIAM Journal on Control and Optimization
Neural Networks: A Comprehensive Foundation

Neural Networks: A Comprehensive Foundation
From Perturbation Analysis to Markov Decision Processes and Reinforcement Learning

Discrete Event Dynamic Systems
Technical Communique: A unified approach to Markov decision problems and performance sensitivity analysis

Automatica (Journal of IFAC)

Quantified Score

Hi-index	0.00

Visualization

Abstract

A policy iteration approach to optimal control problems for a class of nonlinear stochastic dynamic system is introduced. Some parameters and nonlinearities of the system are not required to be known a-priori. An optimality equation is developed based on performance potential. The potential can be estimated by a sample path, and then it is approximated by RBF neural network. As a result, an on-line algorithm is proposed by using a sample path of the given control system.