Neural Networks 63 (2015) 1–9
Contents lists available at ScienceDirect
Neural Networks
journal homepage: www.elsevier.com/locate/neunet
Projective synchronization of fractional-order memristor-based
neural networks
Hai-Bo Bao
a
, Jin-De Cao
b,c,∗
a
School of Mathematics and Statistics, Southwest University, Chongqing 400715, China
b
Department of Mathematics, Southeast University, Nanjing 210096, China
c
Department of Mathematics, Faculty of Science, King Abdulaziz University, Jeddah 21589, Saudi Arabia
a r t i c l e i n f o
Article history:
Received 5 August 2014
Received in revised form 19 September
2014
Accepted 22 October 2014
Available online 31 October 2014
Keywords:
Fractional-order
Memristor-based neural networks
Projective synchronization
Filippov’s solution
a b s t r a c t
This paper investigates the projective synchronization of fractional-order memristor-based neural
networks. Sufficient conditions are derived in the sense of Caputo’s fractional derivation and by combining
a fractional-order differential inequality. Two numerical examples are given to show the effectiveness
of the main results. The results in this paper extend and improve some previous works on the
synchronization of fractional-order neural networks.
© 2014 Elsevier Ltd. All rights reserved.
1. Introduction
During the past few decades, neural networks have been paid
a lot of attention since they have wide applications in the fields
of signal processing, associate memories, combinational optimiza-
tion, automatic control and so on (Cao, Chen, & Li, 2008; Cao &
Wan, 2014; Huang & Feng, 2009; Liang, Wang, Liu, & Liu, 2008;
Lu, Ho, Cao, & Kurths, 2011; Park, Kwon, Lee, Park, & Cha, 2012;
Song, 2009; Wang, Wang, & Liu, 2010; Yang, Zhu, & Huang, 2011).
As a special kind of neural networks, memristor-based neural net-
works have attracted many researchers’ interests. From a system-
theoretic point of view, the memristor-based neural networks
are a class of state-dependent of switching nonlinear systems
which may have undesirable dynamical behaviors such as tran-
sient chaos, oscillators and instability.
Memristor (a contraction for memory resistor) which is consid-
ered to be the fourth passive circuit element was first predicted
by Chua (1971) and was first realized by the Hewlett–Packard
(HP) research team in 2008 (Strukov, Snider, Stewart, & Williams,
2008; Tour & He, 2008). They published their results in Nature.
The new circuit element of memristor shares many properties of
resistor and shares the same unit of measurement (ohm). There
∗
Corresponding author at: Department of Mathematics, Southeast University,
Nanjing 210096, China. Tel.: +86 25 52090596 8535; fax: +86 25 83792316.
E-mail addresses: hbbao07@gmail.com (H.-B. Bao), jdcao@seu.edu.cn
(J.-D. Cao).
are some papers which prove that the memristor exhibits the fea-
ture of pinched hysteresis, the same as the neurons in the human
brain. Because of this feature, more and more researchers con-
struct the memristor-based neural networks to emulate the human
brain through replacing the resistor by the memristor and study
the dynamics of memristor-based neural networks for the purpose
of achieving its successful applications in neural learning, pattern
recognition, associative memories (Chen, Li, Huang, & Wang, 2014;
Hu & Wang, 2010; Li, Hu, & Guo, 2014; Wen, Bao, Zeng, Chen, &
Huang, 2013; Wu, Wen, & Zeng, 2012; Wu & Zeng, 2012; Wu, Zeng,
Zhu, & Zhang, 2011; Yang, Cao, & Yu, 2014; Zhang & Shen, 2013,
2014; Zhang, Shen, & Wang, 2013), etc.
Since Pecora and Carroll (1990) have proposed chaos syn-
chronization, synchronization has become a hot topic in the fields
of neural networks and complex networks. Up to now, there
are many sorts of synchronization, complete synchronization,
anti-synchronization, generalized synchronization, projective syn-
chronization, phase synchronization, lag synchronization, etc. As
pointed out in Wang and He (2008), projective synchronization can
obtain faster communication with its proportional feature, it is sig-
nificant to study projective synchronization of neural networks.
Several different approaches have been used for achieving pro-
jective synchronization, such as linear–nonlinear feedback control
(Chen, Jiao, Wu, & Wang, 2010; Wu & Lu, 2012), adaptive control
(Xiao, Wang, Miao, & Wang, 2012; Zhu, Zhou, Zhou, Wu, & Tong,
2014), impulsive control (Hu, Yang, & Xu, 2008), pinning control
(Hu, Xu, & Yang, 2008; Xiao et al., 2012), and so on.
http://dx.doi.org/10.1016/j.neunet.2014.10.007
0893-6080/© 2014 Elsevier Ltd. All rights reserved.