Please use this identifier to cite or link to this item: http://hdl.handle.net/2440/93639
Citations
Scopus Web of Science® Altmetric
?
?
Type: Conference paper
Title: Neural-network based online policy iteration for continuous-time infinite-horizon optimal control of nonlinear systems
Author: Tang, D.
Chen, L.
Tian, Z.
Citation: Proceedings of the 2015 IEEE China Summit & International Conference on Signal and Information processing, 2015 / pp.792-796
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Issue Date: 2015
ISBN: 9781479919475
Conference Name: 3rd IEEE China Summit & International Conference on Signal and Information processing (ChinaSIP 2015) (12 Jul 2015 - 15 Jul 2015 : Chengdu, China)
Statement of
Responsibility: 
Difan Tang, Lei Chen, and Zhao Feng Tian
Abstract: A new policy-iteration algorithm based on neural networks (NNs) is proposed in this paper to synthesize optimal control laws online for continuous-time nonlinear systems. Latest advances in this field have enabled synchronous policy iteration but require an additional tuning loop or a logic switch mechanism to maintain system stability. A new algorithm is thus derived in this paper to address this limitation. The optimal control law is found by solving the Hamilton-Jacobi- Bellman (HJB) equation for the associated value function via synchronous policy iteration in a critic-actor configuration. As a major contribution, a new form of NN approximation for the value function is proposed, offering the closed-loop system asymptotic stability without additional tuning scheme or logic switch mechanism. As a second contribution, an extended Kalman filter is introduced to estimate the critic NN parameters for fast convergence. The efficacy of the new algorithm is verified by simulations.
Keywords: machine learning; neural network; policy iteration; optimal control; nonlinear system
Description: IEEE Catalog Number: CFP15SIP-USB
Rights: © 2015 IEEE
RMID: 0030032959
DOI: 10.1109/ChinaSIP.2015.7230513
Appears in Collections:Mechanical Engineering publications

Files in This Item:
File Description SizeFormat 
hdl_93639.pdfAccepted version457.16 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.