Further Study on Networked Control Systems with Unreliable Communication Channels
- PDF / 557,277 Bytes
- 10 Pages / 594.77 x 793.026 pts Page_size
- 0 Downloads / 208 Views
ISSN:1598-6446 eISSN:2005-4092 http://www.springer.com/12555
Further Study on Networked Control Systems with Unreliable Communication Channels Xiao Lu, Na Wang, Qingyuan Qi, Xiao Liang*, and Haixia Wang Abstract: This paper focuses on the fundamental problems of linear quadratic gaussian (LQG) control and stabilization problems for networked control systems (NCSs) with unreliable communication channels (UCCs) where packet dropout, input delay and observation delay occur. These basic issues have attracted extensive attentions due to broad applications. Our contributions are as follows. For the finite horizon case, without time-stamping technique, the optimal estimator is derived by using the novelty method of innovation sequences based on the delayed intermittent observations; A necessary and sufficient condition for the optimal control problem is presented on the basis of the solution to the forward and backward difference equations (FBDEs) and two coupled Riccati equations. For the infinite horizon case, it is shown that under certain assumption, the system can stay bounded in the mean square sense if and only if the algebraic Riccati equation admits the unique positive solution. Keywords: Boundedness, LQG control, networked control systems, stabilization.
1.
INTRODUCTION
Nowadays, momentous attention has been paid to networked control systems (NCSs) as they have been playing a significant role in many pivotal infrastructures such as electric system, hydraulic engineering, sewerage system and mobile communication [1–3,20–23,26–28]. Contrasting to classical feedback systems with wired pointto-point link, NCSs have multitude superiority including lower cost of maintenance, higher flexibility and lessened network wiring. The common phenomena lying in unreliable communication channels (UCCs) are packet-dropout and transmission delays which have been studied by many researchers (see [4,5,10,24,25]). However, there still exist many challenging and unsolved issues to be investigated. The study on the case of the packet-dropout can be traced back to [6] and [7], where the linear minimum mean square error estimator (LMMSE) was obtained. Based on the LMMSE, [8] derived the “suboptimal” controller as the well-known separation principle failed due to the multiplicative noise caused by packet-dropout. It is generally known that “intermittent observations” is a pattern of the packet-dropout. With the time-stamping technique, i.e., the packet-dropout process was known, [9] derived the optimal estimator for the system subject to intermittent observations. For the study on the optimal estimation and
control for NCSs without the time-stamping technique, please refer to [11]. Recently, [12] studied the packetdropout scenario, i.e., if the state-packet delivered successfully, the state could be observed precisely. Then the optimal controller and stabilization condition were derived in [12]. However, it should be noted that the state is usually partially observed by sensors in practice, i.e., observation matrices and observation
Data Loading...
