New Trends in Networked Control of Complex Dynamic Systems: Theories and ApplicationsView this Special Issue
Group Synchronization of Nonlinear Complex Dynamics Networks with Sampled Data
Based on a nonlinear consensus protocol, this paper considers the group synchronization of complex dynamical networks with sampled data. Using the Lyapunov method, the group synchronization of the nonlinear complex networks is analyzed. All the nodes in each group can converge to their own synchronous state asymptotically, if the sampled period satisfies some matrix inequality conditions. Furthermore, the theoretical results are verified by some simulations.
Because of the wide application of the complex dynamical networks [1–26], the synchronization problem of the complex networks has become a hot topic recently. A lot of researches concentrate on the continuous information transmission; that is, every node can receive or detect neighbor information of the network all the time [1–3, 5–7, 10]. In , the authors considered the group synchronization of the nonlinear complex dynamical network via pinning control. A new criterion for the cluster synchronization of coupling networks with time-varying delays was established in continuous time in . Moreover, Liu et al.  considered the adaptive synchronization of complex dynamical networks with switching topology by local Lipschitz nonlinearity. In , group consensus of multiagent systems with switching topologies and communication delays was studied in the continuous-time case.
However, in many engineering practices, the information communication between nodes could be interrupted at any time due to some factors, such as the unreliability of communication channels and the limitations of the node detection ability. Hence, it is necessary to consider the information transmission under the discrete state [9, 13–16, 18–21, 23, 26]. Information data on the discrete time can also be regarded as sampled data [11, 25]. In , the group consensus of linear multiagent systems with sampled-data was considered. Xiao and Wang further considered a discrete-time model with time delays in  and analyzed a consensus problem in the existence of the time delay when agents exchanged information between each other. Gao and Wang  studied the continuous-time consensus of multiagent systems with sampled data by time-varying topology. In , consensus of multiple dynamic agents with sampled information was considered. In , average consensus control of networks with sampled data and measurement noises was studied on continuous time. In this paper, we consider the group synchronization of nonlinear complex dynamical network with sampled data.
The rest of this paper is organized as follows. In Section 2, we consider the group synchronization of a complex dynamical network with sampled data and give some assumptions and lemmas. In Section 3, we analyze group synchronization of the proposed network and give the synchronization condition based on linear matrix inequality (LMI). In Section 4, we give the synchronization condition for a special case. In Section 5, the simulations verify the theoretical results. Conclusion is finally summarized in Section 6.
2. Preliminaries and Problem Statement
Consider a complex dynamical network of nodes with sampled data as follows: with where represents the state vectors of the node at time , is continuously differentiable, is coupling strength, and is an inner-coupling matrix, if the node links through its th state with its neighbors ; otherwise . In this protocol, let , and let , where . , , , with . is the neighbors of the node , , where , .
In protocol (2), is the control input: where is an on-off control; if the system is sampling the data, then ; otherwise . If node can get information from node in the same group, then ; otherwise ; if node can get information from node between different groups, then ; otherwise . Thus the coupling configuration matrix can be written as where and represent the coupling configuration of the subgroups, respectively.
Network (1) is said to group synchronization if where the nonlinear function satisfies and , are the synchronous states.
Given a positive real number and a sampled period , we suppose that where shows the discrete time that the node can obtain information from its neighbors and positive integer is a sampled time about the time (), and it satisfies . Under this condition, a linear consensus protocol based on a linear estimation-based sampling period is designed as follows :
Thus, we have
Then the complex dynamical network (1) in every sampled period becomes with
Some assumptions and lemmas are needed.
Lemma 3 (see ). Define , and construct a closed space where and is a constant. For , there exists a constant such that
Lemma 4 (see ). Suppose that and are vectors, and in matrix , the following inequality holds:
In this paper, means that is a positive (or semipositive, negative, or seminegative) definite matrix; , .
3. Group Synchronization Analysis in Complex Network with Sampled Data
In this section, we consider group synchronization problem of the complex networks with sampled data. We have the following theorem.
Theorem 5. For network (1) with protocol (12) of nodes, under Assumptions 1-2 and Lemmas 3-4, if , satisfy then all the nodes in each group can converge to their own synchronous state asymptotically, where
Proof. Construct a Lyapunov function as follows:
Under Assumption 2, we can know Under Assumption 1 and Lemma 3, we can have Then, we can get
Using Lemma 4, we can have Thus, we can obtain If then Therefore, decreases on an interval , , and . From the above discussion, we have
Then, network (1) with sampled data is group-synchronized under protocol (12).
4. A Special Case
In this section, we consider a special case about network (1). For convenience, we let and .
Lemma 6 (see ). If is a symmetric irreducible matrix with and , then for any matrix with all eigenvalues of the matrix are negative.
In networks, when the information transmission is the same between the nodes and , the coupling configuration matrix is symmetric. From Lemma 6, when is symmetric, and are negative. Hence, all of their eigenvalues are strictly negative; one denotes them as Then, using Lemma 6, we can obtain the following result when the coupling configuration matrix is symmetric.
Theorem 7. For network (1) with protocol (12) of nodes, when the coupling configuration matrix is symmetric, under Assumptions 1-2 and Lemmas 3–6, if , satisfy
then all the nodes in each group can converge to their own synchronous state asymptotically, where and .
The proof of Theorem 7 is similar to that of Theorem 5 and here is omitted.
In this section, we give some simulation results of the above discussions. For convenience, let , , , for all , , and , . We consider group synchronization of complex dynamical network with sampled data in the time interval . The curves in the graphs are the locations of nodes and the synchronous targets in the network.
From Figures 1, 2, and 3, the synchronous states are , . From these figures, we can see that all the nodes in each subgroup can converge to their own synchronous targets. But, in Figures 4, 5, and 6, all of the nodes can asymptotically converge to a synchronous state, when the synchronous state is the only one .
In Figures 1 and 4, we simulate synchronization of the network including two subgroups without sampled data. In Figures 2 and 5, we chose as the sampled period. That is, the complex network is sampling at the moment , for . In Figures 3 and 6, we chose as the sampled period; that is, the complex network is sampling at the moment , for .
In these simulation results, we can find that the nodes of the system keep their state value in the first time until before the moment , keeping their state value in time until the moment and so on. By comparing, when choosing the sampled data on every period, the rate of convergence of the nodes in the complex dynamics network is slow. Moreover, the bigger the sampled period is, the slower the rate of convergence of the nodes in the complex dynamics network is.
In this paper, we have investigated the group synchronization problem of a complex dynamical network with sampled data. We prove that the nodes of the network arrive at synchronization in two subgroups if the sampling period satisfies the condition based on the linear matrix inequality (LMI). In addition, we have given some simulation results about the proposed complex network.
Conflict of Interests
The authors declare that there is no conflict of interests regarding the publication of this paper.
This work was supported by the National Natural Science Foundation of China under Grant no. 61304049, Funding Project for Academic Human Resources Development in Institutions of Higher Learning under the Jurisdiction of Beijing Municipality (PHR201108055), and Science and Technology Development Plan Project of Beijing Education Commission (nos. KM201310009011 and KM201310009013).
B. Liu, P. F. Wei, X. F. Wang et al., “Group synchronization of complex networks with nonlinear dynamics via pinning control,” in Proceeding of the 32nd Chinese Control Conference, pp. 235–240, 2013.View at: Google Scholar
H. Zhang, H. C. Yan, F. W. Yang, and Q. J. Chen, “Distributed average filtering for sensor networks with sensor saturation,” IET Control Theory and Applications, vol. 7, no. 6, pp. 887–893, 2013.View at: Google Scholar
H. C. Yan, Z. Z. Su, H. Zhang, and F. W. Yang, “Observer-based control for discrete-time stochastic systems with quantization and random communication delays,” IET Control Theory and Applications, vol. 7, no. 3, pp. 372–379, 2013.View at: Google Scholar
Y. J. Yu, M. Yu, J. P. Hu, and B. Liu, “Group consensus of multi-agent systems with sampled data,” in Proceeding of the 32nd Chinese Control Conference, pp. 7168–7172, 2013.View at: Google Scholar
H. C. Yan, H. B. Shi, H. Zhang, and F. W. Yang, “Quantized control for networked delayed systems with communication constraints,” Asian Journal of Control, vol. 15, no. 5, pp. 1468–1476, 2013.View at: Google Scholar
L. X. Zhang, H. J. Gao, and O. Kaynak, “Network-induced constraints in networked control systems—a survey,” IEEE Transactions on Industrial Informatics, vol. 9, no. 1, pp. 403–416, 2013.View at: Google Scholar