Reinforcement Learning NN-based Controller Design for Aero-engine

DOI

10.2991/iccse-15.2015.24How to use a DOI?

Keywords

Aero-engine, Neural Networks, Reinforcement Learning, Cost function

Abstract

A reinforcement learning NN-based controller for aero-engine is presented, which is suitable for tracking problems at a steady-state operating point. The proposed controller design has two entities:an action NN that is designed to produce optimal signal for aero-engine and a critic NN that approximates certain strategic utility function which evaluates the performance of the action network . In accordance with a turbofan engine, the control system is designed at the selected operating point. The simulation results show that the perfect performance of the controller and the controller has strong anti-interference ability.

Copyright

© 2015, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Hong-Mei Zhang
AU  - Yu-Ling Liang
AU  - Guang-Yan Xu
PY  - 2015/07
DA  - 2015/07
TI  - Reinforcement Learning NN-based Controller Design for Aero-engine
BT  - Proceedings of the 2015 International Conference on Computational Science and Engineering
PB  - Atlantis Press
SP  - 138
EP  - 145
SN  - 2352-538X
UR  - https://doi.org/10.2991/iccse-15.2015.24
DO  - 10.2991/iccse-15.2015.24
ID  - Zhang2015/07
ER  -