Online learning optimal control for decentralized stabilization of nonlinear interconnected systems

Author

Derong Liu ; Ding Wang ; Hongliang Li

Author_Institution

State Key Lab. of Manage. & Control for Complex Syst., Inst. of Autom., Beijing, China

fYear

2013

fDate

26-29 May 2013

Firstpage

229

Lastpage

234

Abstract

In this paper, the decentralized stabilization of a class of nonlinear interconnected systems is investigated using online learning optimal control approach. By introducing cost functions that reflect the bounds of interconnections, the optimal controllers of the isolated subsystems can be designed first. Then, it is proved that the decentralized control strategy of the overall system can be developed based on the optimal control policies. Next, the online policy iteration algorithm is presented to solve the Hamilton-Jacobi-Bellman equations. The cost functions and control policies are obtained approximately by employing neural networks. In addition, a simulation example is provided to verify the effectiveness of the present decentralized control strategy.

Keywords

decentralised control; learning (artificial intelligence); neurocontrollers; nonlinear systems; optimal control; stability; Hamilton-Jacobi-Bellman equations; cost functions; decentralized control; decentralized stabilization; neural networks; nonlinear interconnected systems; online learning optimal control policies; online policy iteration algorithm; optimal controllers; Algorithm design and analysis; Approximation algorithms; Cost function; Decentralized control; Equations; Interconnected systems; Optimal control;

fLanguage

English

Publisher

ieee

Conference_Titel

Cyber Technology in Automation, Control and Intelligent Systems (CYBER), 2013 IEEE 3rd Annual International Conference on

Conference_Location

Nanjing

Print_ISBN

978-1-4799-0610-9

Type

conf

DOI

10.1109/CYBER.2013.6705450

Filename

6705450