Paper Info
Title
On LQG control design for network systems with/without acknowledgments using a particle filtering technology.
Abstract
This paper focuses on the behavior of a linear quadratic Gaussian (LQG) controller for discrete-time systems utilizing industrial communication protocols. A new design of LQG problem is investigated for constrained networked control systems by using a new class of a quadratic cost function with a communication cost. This paper is considered two industrial communication protocols namely: the transportation control protocol (TCP), in which the sender should receive an acknowledgment regarding the delivery of the sending packets and the user datagram protocol (UDP) where the acknowledgment signal is absence. In the latter one, the driver (actuation) does not afford acknowledgment to inform the controller/estimator regarding the reception of the action packets. In such setting, the UDP-like protocol can react nonlinearly, where the absence of the acknowledgment creates a nonlinear optimization control law. In this sense, a LQG with standard Kalman filtering is failed because the separation principle can not reached. Under this setting, a suboptimal methodology can be successfully implemented to compensate the absence of communication and acknowledgment packets. The prescribed performance of networked system is sustained in the presence of communication constraints. The probabilities of the data signal dropouts are expressed in terms of random Bernoulli processes. Then, simulation results are provided to demonstrate the validity of the proposed schemes and to compare them with others from literature.
Year
DOI
Venue
2019
10.1016/j.amc.2019.04.012
Applied Mathematics and Computation
Keywords
Field
DocType
LQG control design,Networked control systems,Communication cost,Network protocols,Packet dropouts
Mathematical optimization,Control theory,User Datagram Protocol,Linear-quadratic-Gaussian control,Separation principle,Control theory,Nonlinear programming,Network packet,Kalman filter,Control system,Mathematics
Journal
Volume
ISSN
Citations 
359
0096-3003
0
PageRank 
References 
Authors
0.34
0
2
Name
Order
Citations
PageRank
Nezar M. Alyazidi122.07
Magdi S. Mahmoud279098.50