第
27
卷第
2
期
2010
年
2
月
控制理论与应用
Control
Theory
& Applications
Vo
l.
27
No.
2
Feb.20
1O
Article
ID:
1000-8152(2010)02-0181-07
Smart
variable
structure
control
of
complex
network
with
time-varying inner-coupling
matrix
to its equilibrium
YANG
Yue-quan\
YU
Xing-huo
2
,
ZHANG
Tian-ping
1
(1.
Department
of
Automation
,
College
of
Information
Engineering
,
Yangzhou
University
,
Yangzhou
Jiangsu
225009
,
China;
2.
Platform
Technologies
Research
Institute
,
RMIT
University
,
Melbourne
VIC
3001
,
Australia)
Abstract: The novelty
in
this
pap
巳
r
lies in the
establishm
巳
nt
of smart controller and suitable multiple sliding mode
manifolds according
to
node chaos dynamics
of
complex networks with time-varying
inn
巳
r-coupling
configuration. The
smart variable structure control
for
asymptotical synchronization
to
its equilibrium
is
d
巳
veloped
based on
th
巳巳
rgodicity
characteristic of chaos nodes, without the
involv
巳
ment
of linearization and
oth
巳
r
ideal assumptions. The scheme enables the
behavior of complex networks
to
approach the desired manifolds,
and
eventually realizes the asymptotical synchronization.
Finally
, the simulations based on the Lorenz chaos complex network under three topological configurations further verify
the robustness
and
巳征
'ectiveness
of
the proposed
schem
巳.
Key words: complex network; synchronization; smart control; variable structure control
CLC
number: TP273 Document code: A
时变内藕合复杂网络的平衡态同步
smart
变结构控制
杨月全
1
余星火
2
张天平
1
(1.扬州大学信息工程学院臼动化专业部,江苏扬州
225009;
2.
皇家墨尔本理工大学工程技术平台研究院,澳大利亚维多利亚州
3001)
摘要:当前同步控制问题是复杂网络研究的热点之一.本文针对具有时变内祸合结构的复杂网络,利用结点混
沌动态的各态历经性,通过构造合适的滑模面,提出了
smart
变结构控制器的设计策略.该策略可使复杂网络动态
行为趋向于所构造的全局吸引区域,从而最终实现复杂网络在平衡态的渐近同步.最后,基于
3
种不同拓扑结构
的
Lorenz
结点动态的复杂网络进行仿真实验表明该控制方案具有较好的鲁棒性和有效性.
关键词:复杂网络,同步;
smart
控制;变结构控制
1
In
troduction
nodes
are
inter-connected
by
directed
or
undirected
To
gain
an
insight
into
the
mechanism
of
complex
edges
or
links
with
di
旺
érent
topological
structures.
Cur-
network
operation
and
even
control
and
prediction
of
network
behavior
,
complex
network
systems
have
been
extensively
studied
recently.
Increasing
research
atten-
tion
has
been
drawn
to
the
control
and
analysis
of
com-
plex
network
systems.
Complex
network
systems
are
ubiquitous
,
including
many
natura1
or
man-made
sys-
tems
,
such
as
socia1
network
systems
, neura1
network
systems
,
the
Intemet
, 10gistic
network
systems
, e1ectri-
cal
power
grids
,
satellite
network
guidance
systems
,
and
so
on.
In
genera1,
network
systems
can
be
represented
by
means
of
graphs
in
mathematica1
terms
where
many
Receiv
巳
d
date:
2009-06-20;
Revised
dat
巳:
2009-10-02.
rently
,
main
comp1ex
networks
models
are
regu1ar
net-
works
mode1,
random
network
mode1,
small-world
net-
work
mode1,
and
sca1e
free
network
model
[1~81.
Collective
motions
of
comp1ex
networks
have
been
the
subject
of
considerable
interest
within
the
science
and
techno1ogy
communities
over
1ast
decade.
Espe-
cially
,
one
of
the
interesting
and
significant
phenomena
in
comp1ex
networks
is
the
synchronization.
Recently
,
synchronization
research
of
comp1ex
networks
has
been
reported
in
the
1iterature[9~211.
The
study
of
the
syn-
chronization
in
a sca1e-
free
dynamica1
network
has
been
Foundation
item:
supported
by
the
National
Natural
Science
Foundation
of
China(60774017
,
60874045);
the
Open
Projects
of
Key
Laboratory
of
Complex
Systems
and
Inte
l1i
gence
Science
of
Chin
巳
se
Academy
of
Sciences(20060101).