ORIGINAL ARTICLE
Minimum entropy control of nonlinear ARMA systems
over a communication network
Jianhua Zhang Æ Hong Wang
Received: 27 January 2007 / Accepted: 15 June 2007
Springer-Verlag London Limited 2007
Abstract In this paper, the entropy concept has been
utilized to characterize the uncertainty of the tracking error
for nonlinear ARMA stochastic systems over a communi-
cation network, where time delays in the communication
channels are of random nature. A recursive optimization
solution has been developed. In addition, an alternative
algorithm is also proposed based on the probability density
function of the tracking error, which is estimated by a
neural network. Finally, a simulation example is given to
illustrate the efficiency and feasibility of the proposed
approach.
Keywords Networked control system Entropy
Neural networks
1 Introduction
Major advancements in the area of communication and
computer networks have made it possible to insert them in
feedback systems in order to achieve real time require-
ments. Such networked control systems (NCSs) have been
used in various areas, such as: automotive industry, tele-
autonomy, teleoperation of robots and automated manu-
facturing systems. However, control over networks makes
the analysis and design of the closed-loop systems com-
plex.
A major problem in NCSs is that the network introduces
delays. The induced delay is inevitable and typically
deteriorates the NCS’s performance. As such, it is impor-
tant to develop control strategies that can cope with the
closed loop control when the actuator and sensor channels
are subjected to time delays. In this context, some meth-
odologies have been developed to deal with network de-
lays. In the past years, the approaches for linear time-delay
systems have been presented [1] and extended to NCSs [2,
3]. An augmented state vector method [4] is proposed to
control a linear system over a periodic delay network.
Queuing mechanisms [5, 6] are developed, which utilize
some deterministic or probabilistic information of NCSs
for control purposes. Intelligent control method [7], such as
fuzzy logic and genetic algorithm, is used to analyze and
design NCSs. Model-based method [8] is proposed to
model NCSs to be a continuous system with fixed sampling
period. In addition, stochastic control method is also
proposed, linear quadratic control systems with random
network delays is studied in [9, 10]. In [11], NCSs are
modeled to be a discrete-time Markovian jump linear
system with mode-dependent time delays.
The delay characteristics on NCSs basically depend on
the type of a network used. Delays over random access
networks are stochastic in nature, such as CAN and
Ethernet. The significant parts of random network delays
are the waiting time delays due to queuing and frame
collision on the networks. When an NCS operates across
networks, several other factors can increase the random-
ness on network delays such as the queuing time delays at a
switch or a router, and the propagation time delays from
different network paths. In addition, a cyclic service net-
work connected to a random access network also leads to
J. Zhang (&)
Department of Automation,
North China Electric Power University,
Beijing 102206, China
e-mail: zhangwu@public.bta.net.cn
H. Wang
Control Systems Centre, The University of Manchester,
P.O. Box 88, Manchester, UK
e-mail: hong.wang@manchester.ac.uk
123
Neural Comput & Applic
DOI 10.1007/s00521-007-0140-8