Title :
Structure preserving reduction of port hamiltonian system using a modified LQG method
Author :
Wu Yongxin ; Hamroun, Boussad ; Le Gorrec, Yann ; Maschke, Bernhard
Author_Institution :
LAGEP, Univ. de Lyon, Lyon, France
Abstract :
This paper proposes a controller reduction method for the port Hamiltonian system by using a modified LQG method. We first use the LQG method to design two passive type controllers which are equivalent to the control of port Hamiltonian system by interconnection. One of these LQG type method permit us to define a LQG balanced realization by computing its LQG Grammians. Then we use the effort-constraint method to achieve a reduced order port Hamiltonian system and design a reduced order passive type LQG controller. Finally, the method is illustrated on a mass-spring system by numerical simulations of the closed loop system with the full order passive LQG controller and with its reduced order controller.
Keywords :
closed loop systems; linear quadratic Gaussian control; nonlinear control systems; numerical analysis; reduced order systems; LQG Grammians; LQG balanced realization; closed loop system; controller reduction method; effort-constraint method; full order passive LQG controller; linear quadratic Gaussian method; mass-spring system; modified LQG method; numerical simulations; port Hamiltonian system; reduced order controller; structure preserving reduction; Covariance matrices; Matrix decomposition; Ports (Computers); Riccati equations; Standards; Symmetric matrices; LQG method; controller reduction; model reduction; passive controller; port Hamiltonian system;
Conference_Titel :
Control Conference (CCC), 2014 33rd Chinese
Conference_Location :
Nanjing
DOI :
10.1109/ChiCC.2014.6895525