A generalized policy iteration adaptive dynamic programming algorithm for optimal control of discrete-time nonlinear systems with actuator saturation

Qiao Lin, Qinglai Wei, Bo Zhao

Research output: Chapter in Book/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

In this study, a nonquadratic performance function is introduced to overcome the saturation nonlinearity in actuators. Then a novel solution, generalized policy iteration adaptive dynamic programming algorithm, is applied to deal with the problem of optimal control. To achieve this goal, we use two neural networks to approximate control vectors and performance index function. Finally, this paper focuses on an example simulated on Matlab, which verifies the excellent convergence of the mentioned algorithm and feasibility of this scheme.

Original languageEnglish
Title of host publicationAdvances in Neural Networks - ISNN 2017 - 14th International Symposium, ISNN 2017, Proceedings
EditorsFengyu Cong, Qinglai Wei, Andrew Leung
PublisherSpringer Verlag
Pages60-65
Number of pages6
ISBN (Print)9783319590806
DOIs
Publication statusPublished - 2017
Externally publishedYes
Event14th International Symposium on Neural Networks, ISNN 2017 - Sapporo, Hakodate, and Muroran, Hokkaido, Japan
Duration: 21 Jun 201726 Jun 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10262 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference14th International Symposium on Neural Networks, ISNN 2017
Country/TerritoryJapan
CitySapporo, Hakodate, and Muroran, Hokkaido
Period21/06/1726/06/17

Keywords

  • Adaptive dynamic programming
  • Neural network
  • Optimal control
  • Saturating actuators

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science

Fingerprint

Dive into the research topics of 'A generalized policy iteration adaptive dynamic programming algorithm for optimal control of discrete-time nonlinear systems with actuator saturation'. Together they form a unique fingerprint.

Cite this