Abstract
In this paper, exponential synchronization problem of uncertain complex dynamic networks with time delay is studied via adaptive event-triggered control. Considering the influence of external environment, a new dynamic event-triggered mechanism is proposed, in order to reduce the transmission signal among nodes and reduce the consumption of communication resources. Moreover, in the proposed control mechanism, the controller is adaptive, that is, it only works when the triggering conditions are satisfied. Then, according to the designed adaptive event-triggered control strategy, the sufficient conditions for exponential synchronization are obtained by using Lyapunov functions and inequality technique. In addition, it is proved that the system can avoid Zeno behavior. At last, using two examples to verify the feasibility of the results.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
Complex dynamic networks (CDNs) are networks consisted of a great deal of nodes in a specific topology, which are widely used in a large amount of real systems, such as metabolic networks, mobile communication networks, urban transportation networks, neural networks etc [1,2,3,4,5,6]. As a result, CDNs have become an attractive research topic in practical engineering applications.
CDNs have many interesting dynamic behaviors, such as space-time chaos, spiral waves, and synchronization. In particular, synchronization, as a collective behavior of CDNs, has attracted more and more scholars to explore and discuss in the field of physics, biology, and other research [7,8,9]. Therefore, in recent years, plenty of research has been done on various analyses of network synchronization, including dissipative synchronization, exponential synchronization, asymptotic synchronization, finite-time synchronization, etc. In literature [10], the sliding mode control method was used to study the asymptotic synchronization of time-delayed CDN. [11] researched how CDN achieve exponential synchronization under feedback control and periodic intermittent noise. The literature [12] demonstrated that the CDN achieves finite-time synchronization problem under the influence of sampled-data control. Among them, several studies have been conducted on the exponential synchronization problem of CDNs [13, 14].
Through the study of CDN, scholars have found that CDN is not a fixed structure and is vulnerable to various disturbances from the external environment. Therefore, in order to overcome this problem, various control strategies are proposed to synchronize it, such as pulse control, sliding mode control, sampled-data control, event-triggered control, etc. Among them, literature [15] studied the global asymptotic synchronization of nonlinear complex neural networks under the control of event-triggered time delay pulses. Hu and others in the literature [16] studied CDN quasi-synchronization of event-triggered control in a time-varying topology framework. In [17], authors studied the global \({H_\infty }\) synchronization of a class of aperiodic CDNs by sampled-data control. In literature [18], global asymptotic synchronization of non-equivalent CDNs with time delay and unknown perturbations was studied by sliding mode control. However, in the normal operation of CDN, a large number of data packets are sent to the communication channels, which consumes a lot of communication resources, so an event-triggered control (ETC) strategy is proposed.
In recent decades, the ETC strategy is a control strategy for data transmission according to the triggering conditions, that can only be transmitted when the conditions of event-triggered are met, so this control strategy has been extensively explored and researched by scholars. This method was first put forward in the literature [19] for the study of system stability. The research method is to transmit the data to the actuator to complete the control task when the error signal violates the given triggering condition. It is found that the ETC strategy can efficiently reduce the number of samples, save communication resources and reduce communication burden. Therefore, researchers have increasingly focused on the synchronization of CDN via the ETC strategy, and have obtained a number of interesting findings [20,21,22,23,24,25,26,27,28]. However, in much literature, the study of ETC is static, that is, triggering conditions are set in advance [29, 30]. In addition, in the existing literature, the authors combined ETC with adaptive control to consider system consistency, such as [31]. In fact, in recent decades, adaptive control strategy has received more and more attention as a control strategy that adapts to the evolution results of the controlled network [32, 33]. As is known to all, in the actual network, the internal system of the CDN is inevitably interfered by the external environment and therefore has a certain uncertainty [34, 35]. Besides, time delay is inevitable [36, 37]. In addition, because the ETC has reduced the number of sampling, so it makes the controller more flexible under adaptive control.
In view of the above discussion, this paper proposes a dynamic adaptive ETC scheme, and studies the exponential synchronization of CDN with uncertain time delay by designing a suitable adaptive ETC control scheme. At the same time, according to the matrix inequality technique, the conditions for exponential synchronization are obtained. The main features are expressed as follows:
-
1.
Due to the sensitivity of the internal system to the external environment, CDN with delay uncertainty is considered, which is more common. According to the designed adaptive law, the event-triggered conditions can change dynamically.
-
2.
Putting forward a new dynamic ETC, and the results of existing literature [38,39,40] are compared with the method proposed in this paper, which avoids the tedious transmission signal, reduces the number of sampling, and effectively saves communication resources.
-
3.
According to the stability of Lyapunov, the conditions that there is no Zeno behavior are obtained, the Zeno behavior is avoided, and the exponential synchronization with time delay CDN is realized.
The rest of the article is arranged below. The second part designs an ETC control strategy and gives some assumptions, definition, and lemma. The third part combines the previous knowledge to draw the main conclusions. The fourth part verifies the advantages of the designed ETC through two simulation examples. Finally, the conclusion is given.
2 Problem Formulation and Preliminaries
In the paper, the continuous-time CDN with time delays containing M coupled nodes is taken into account below:
where \(i = 1,\;2, \ldots ,\;M;\;\) \({x_i}(t) \in {R^n}\) denotes the state variable of the i th dynamic node; \({u_i}\left( t \right) \in {R^n}\) is the control input of the i th node; as well as \(\Gamma \in {R^{n \times n}}\) is a known matrix denoting the inner-coupling. \(\tau (t)\) is time delay, and \(\tau (t) \in (0\,,\tau )\), \(l({x_i}(t - \tau (t)))\) is a nonlinear function. \(A \in {R^{n \times n}}\) is a known matrix.
Assumption 1
In the convention, the weighted matrix \(S = {({s_{ij}})_{M \times M}}\), satisfies that all elements are non-negative that \({s_{ij}} \ge 0\), \(i \ne j\), otherwise \({s_{ij}} = 0\). Moreover, \({s_{ii}} = - \sum \limits _{j = 1,j \ne i}^M {{s_{ij}}} \).
The objective of this article is to synchronize the CDN. Let the synchronization error vector be \({v_i}(t) = {x_i}(t) - {\bar{v}}(t)\), here, the corresponding response system \({\bar{v}}(t)\) is proposed, which stands for the state trajectory of the isolated node, and the condition \(\dot{\bar{ v}}(t) = Av(t) + l({\bar{v}}(t - \tau (t)))\) is met. Therefore, the error system can be represented below:
where \({\tilde{l}}({v_i}(t - \tau (t))) = l({x_i}(t - \tau (t))) - l({{\bar{v}}_i}(t - \tau (t)))\).
It is worth noting that in practical applications, the communication among nodes will be affected. To this end, an event-triggered strategy is proposed to determine whether the current signal should be sent to reduce the waste of communication resources.
Remark 1
Although the existing literatures have proposed to adopt the ETC strategy to synchronize CDNs, most of them are static and need to exchange information with adjacent nodes, which may waste resources and cause information blockage. Therefore, this paper proposes dynamic ETC strategy, which greatly reduces the sampling times and saves resources.
In addition, the triggering time of the basic condition is defined as the following formula:
where \({\sigma _i} \in (0,1)\), and \({\varsigma _i}(t)\;(i = 1,\;2,\; \ldots ,\;M)\) mean a dynamic parameter and meet the following relation:
and \(\alpha > 0\).
Besides, parameters \(\alpha \) and \(\sigma _i\) indicate the tightness of the triggering process and the attenuation rate of the filter, respectively.
Remark 2
Furthermore, the event-triggered mechanism in this paper has a dynamic entry, that is \({\varsigma _i}(t)\), whereas [31] has only a static entry. In addition, the event-triggered control policy only uses its own state and is completed if and only if the triggering condition is true. Therefore, dynamic adaptive ETC can reduce the number of triggers more effectively.
Remark 3
Zeno behavior refers to the control that is triggered an infinite number of times in a finite period of time under the event-triggered control. And in the subsequent proof of Theorem 2, it can be seen that Zeno behavior is avoided.
According to the above discussion, the controller designed is expressed as follows:
Remark 4
The significance of considering sign function in control design is to ensure the stability and reliability of the system. In this paper, the sign function is used to judge whether a variable reaches a certain threshold value, so as to trigger the corresponding event or control strategy.
Here, \({K_i}\) represents the gain matrix, and \({\mu _i}(t)\) stands for adaptive strength, and satisfies the expression:
Remark 5
In order to study exponential synchronization of CDN, adaptive ETC is proposed in this paper. This strategy is dynamic and introduces adaptive intensity. Furthermore, in contrast with the existing results in [41, 42], the designed controller is more general.
Substituting the above Eq. (4) into Eq. (2), we get the following equation:
Remark 6
The Kronecker product is an operation between two matrices of any size, and it is a special form of tensor basis. Given two matrices \(A \in {R^{m \times n}}\) and \(B \in {R^{p \times q}}\), then the Kronecker product of these two matrices is a partitioned matrix of space \({R^{mp \times nq}}\)
Using the Kronecker product, we can simplify the above equation to:
Definition 1
[43] If the error system (5) is exponentially stable and satisfies the relation:
where \(\beta > 0\), \(\alpha > 0\), then the CDN is exponential synchronization.
Assumption 2
In the nonlinear function \(l( \cdot )\), for any \({f_1}\), \({f_2}\), there exists a constant \(l > 0\), so that the relation holds: \(||l({f_1}) - l({f_2})|| \le l||{f_1} - {f_2}||\).
Lemma 1
[44] Given two vectors \(a,b \in {R^n}\), any constant \(\eta > 0\) and a positive definite matrix \(W \in {R^{n \times n}}\), the following inequality holds:
3 Main Results
This part is on account of dynamic event-triggered mechanism, the stability of error system is analyzed. First of all, the sufficient condition of exponential synchronization is explored based on Lyapunov function theory. Then it is clear that the CDN can avoid Zeno behavior via the designed dynamic event-triggered controller.
Set
Theorem 1
For the given parameters \(\beta ,\,\sigma ,\,l > 0\) and matrices \(A,\;S,\;\Gamma \in {R^{n \times n}}\), if there is matrix \(G \in {R^{n \times n}}>0\) and positive number \(\lambda \), satisfying inequality below:
where \({W_{11}} = diag( {-\lambda _1}{-\alpha _1}+ 1,\,{-\lambda _2}{-\alpha _2}+ 1,\, \ldots ,\,{-\lambda _M}{-\alpha _M}+ 1)\),
\(K = diag\{ {K_1},\;{K_2},\; \ldots ,\;{K_M}\} \),
then the uncertain CDN with time delay can achieve exponential synchronization under ETC.
Proof
The Lyapunov function is constructed below:
where
Deriving the above equations, it can be obtained that:
Using Assumption 2 and Lemma 1, we can scale the inequality to:
By using the equation \({v_i}(t) = {v_i}(t_k^i) - {\theta _i}(t)\), the formula can be obtained:
and
By combining Eqs. (12–14) to (9), it is available that:
In addition, we can also obtain:
Moreover, according to the triggering condition (3), the formula can easily obtained:
By summarizing the inequalities (8–17), we can get the following inequality:
According to the conditions known by the Theorem 1, the formula can be get: \(\dot{V}(t) - \lambda V(t) < 0\).
And, it is not hard to figure out that: \(V(t) - V(0){e^{ - \lambda t}} \le 0\).
Thus, there is:
Also known by \(||v(t)|{|^2} = {V_1}(t) < V(t) \le (||v(0)|{|^2} + \sum \limits _{i = 1}^M {{{\varsigma _i}(0)}} ){e^{ - \lambda t}}\).
Namely
Based on the concept of exponential synchronization given above, we can draw a conclusion that the system (2) realizes exponential synchronization. \(\square \)
Remark 7
From the above conclusion, it can be concluded that the error system v(t) is exponentially convergent, namely \(||v(t)|| \le \sqrt{||v(0)|{|^2} + \sum \limits _{i = 1}^M {{{\varsigma _i}}(0)} } {e^{ - \frac{{\lambda t}}{2}}}\). And its norm is affected by M and \(\lambda \).
Theorem 2
After the whole of conditions in Theorem 1 are met, the system under ETC will avoid the Zeno behavior.
Proof
First, according to equation \({\theta _i}(t) = {v_i}(t_k^i) - {v_i}(t)\), it can receive that:
Use the proof by contradiction.
Suppose that the system has Zeno behavior, namely there is a time \(\Delta T_i^*\) which makes \(\mathop {\lim }\limits _{k \rightarrow \infty } (t_{k + 1}^i - t_k^i) = \Delta T_i^*\) as well as \(\mathop {\lim }\limits _{k \rightarrow \infty } t_k^i = \Delta T_i^*\) true.
In addition, the condition \({[l(x) - l(y) - M(x - y)]^{\textrm{T}}}[l(x) - l(y) - N(x - y)] \le 0\) satisfied by the nonlinear function \({\tilde{l}}( \cdot )\) gives the inequality below:
where \(\gamma = \frac{1}{2}[||M + N|| + \sqrt{{\lambda _{\max }}{{(M - N)}^{\textrm{T}}}(M - N)} ]\), \({\bar{w}} = \sqrt{||v(0)|{|^2} + \sum \limits _{i = 1}^M {{{\varsigma _i(0)}} }} {e^{ - \frac{{\lambda t}}{2}}}\).
Moreover, there exists a number \({\varphi _i>0}\), makes \(||{u_i}(t)|| \le {\varphi _i}\), \(||{v_j}(t)|| \le {\varphi _i}\).
Therefore, combining the above inequalities, it can be obtained:
where \({{\bar{\varphi }} _i} = \gamma {\bar{w}} + {\bar{s}}||\Gamma ||\sum \limits _{j = 1}^M {{v_j}(t)} + {\varphi _i}\).
Integral the above equation to get:
The event-triggered mechanism shows that, when the sufficient condition of the triggering condition is not met, it can be obtained:
here, \({\varphi _i} = \sqrt{{\varsigma _i}(0) + {\sigma _i}||{v_i}(0)|{|^2}} \).
Therefore, we can obtain the inequality as follows:
From this we can easily derive inequalities:
Make \({\kappa _0} = \frac{1}{{||A||}}\ln \left( \frac{{{\varphi _i}}}{{||{v_i}(t_k^i)|| + \frac{{{{{\bar{\varphi }} }_i}}}{{||A||}}}}{e^{ - ({\lambda _i} + 1)\frac{{{T_0}}}{2}}} + 1 \right) > 0\), that is, there is a large enough positive integer \({N_0}\), making \(t_k^i \in [{T_0} - {\kappa _0},\,{T_0}],\quad \forall k \ge {N_0}\) is established.
Assume that \(t_k^i\) is on behalf of the k th triggering time of node \({v_i}\). Then, set \(t_{k + 1}^i\) and \({\tilde{t}}_{k + 1}^i\) denote the next triggering moment decided by the triggering condition (3) and Eq. (27) above, respectively.
As a result, it can be easily obtained that:
This result is inconsistent with the result of equation (27), and the hypothesis is not valid. Therefore, Zeno behavior can be avoided.
In addition, without considering nonlinear CDNs, system (2) can be converted to:
\(\square \)
Corollary 1
For the given parameters \(\beta ,\,\sigma > 0\) and matrices \(A,\;S,\;\Gamma \in {R^{n \times n}}\), if there is a positive number \(\lambda \), the following matrix is true:
where \({{\bar{W}}_{11}} = diag( - {\alpha _1} + {\lambda _1},\, - {\alpha _2} + {\lambda _2},\, \ldots \, - {\alpha _M} + {\lambda _M}) \),
\(K = diag\{ {K_1},\;{K_2},\; \ldots ,\;{K_M}\} \),
then the CDN can achieve exponential synchronization under ETC. As well as the Zeno behavior can be avoided.
Proof
First, construct the Lyapunov function below:
Here, \({V_1}(t) = \sum \limits _{i = 1}^M {v_i^{\textrm{T}}(t)} {v_i}(t)\), \({V_2}(t) = \sum \limits _{i = 1}^M {{{{{\bar{\varsigma }} }_i}(t)}} \).
The derivative of V(t) can be obtained from the above equation:
From Lemma 1 and Eq. (eq4), it can be easily obtained:
Thus, it can be get:
Combining the (31), (33), (35), the inequality can be obtained below:
where \({\bar{\delta }} (t) = {\left[ {\begin{array}{*{20}{c}} {{v^{\textrm{T}}}(t)}&{{\theta ^{\textrm{T}}}(t)}&{{{{\bar{\xi }} }^{\textrm{T}}}(t)} \end{array}} \right] ^{\textrm{T}}}\), \({\bar{\varsigma }} (t) = {\left[ {\begin{array}{*{20}{c}} {{\bar{\varsigma }} _1^{\frac{1}{2}}(t)}&{{\bar{\varsigma }} _2^{\frac{1}{2}}(t)}&\cdots&{{\bar{\varsigma }} _M^{\frac{1}{2}}(t)} \end{array}} \right] ^{\textrm{T}}}\).
According to Schur’s lemma, it is easy to get \(\dot{V}(t) - \lambda V(t) < 0\).
Integrating the above equation can get:
Because of \(||v(t)|{|^2} \le V(t)\), it can be reached that:
It can be seen that CDN achieves exponential synchronization. Furthermore, by referring to the proof of Theorem 2, it can be obtained that Zeno behavior in CDN (29) can be avoided. \(\square \)
Remark 8
Corollary 1 studies exponential synchronization in linear and delay-free CDNs, making the results more general.
4 Numerical Examples
We demonstrate the superiority of the proposed control method through two examples.
Example 1
Here an uncertain CDN with time delay with three nodes and two topologies is considered, the system is expressed as follows:
In the above equation, let \(A = \left[ {\begin{array}{*{20}{c}} { - 3}&{}0&{}0&{}0&{}0&{}0\\ 0&{}{ - 3}&{}0&{}0&{}0&{}0\\ 0&{}0&{}{ - 3}&{}0&{}0&{}0\\ 0&{}0&{}0&{}{ - 3}&{}0&{}0\\ 0&{}0&{}0&{}0&{}{ - 3}&{}0\\ 0&{}0&{}0&{}0&{}0&{}{ - 3} \end{array}} \right] \), \(\Gamma = \left[ {\begin{array}{*{20}{c}} 0.5&{}0\\ 0&{}0.5 \end{array}} \right] \), \(S = \left[ {\begin{array}{*{20}{c}} { - 1}&{}0&{}1\\ 2&{}{ - 2}&{}0\\ 0&{}2&{}{ - 2} \end{array}} \right] \).
In addition, the nonlinear function is chosen as:
And based on the Assumptions and Lemma 1 given above, the value of the parameters can be chosen as: \(\alpha \mathrm{{ = }}3\), \(\sigma \mathrm{{ = }}0.8\), \(\lambda \mathrm{{ = }}1.5\), \(l = 2\).
Moreover, we select the initial values of the system state below: \({v_1}(0) = {\left[ {\begin{array}{*{20}{c}} 2&{ - 2.5} \end{array}} \right] ^{\textrm{T}}}\), \({v_2}(0) = {\left[ {\begin{array}{*{20}{c}} {1.5}&{ - 1} \end{array}} \right] ^{\textrm{T}}}\), \({v_3}(0) = {\left[ {\begin{array}{*{20}{c}} {3.5}&{ - 2} \end{array}} \right] ^{\textrm{T}}}\). And by using LMI to calculate the matrix in Theorem 1, the control gain matrices can be expressed as follows:
Besides, parameters in the controller and system can be set to \(\gamma = 1\), \(\alpha = 3\).
Figures 1, 2, 3 and 4 show the simulation results for the sake of demonstrating the superiority of the results obtained. Figure 1 indicates the state dynamic trajectory of each node in the CDN system. Figure 2 displays the state trajectories of the synchronization errors, which indicates convergence to a range. Figures 3 and 4 show the dynamic trajectories of adjective law and adaptive strength respectively. As a consequence, from above four figs, it is observed that each node of the CDN under the control of the deaigned ETC can realize synchronization. Figure 5 demonstrates that without a controller, the system is not stable. Not only that, Fig. 6 shows the event interval of each node, which greatly reducing the number of information transfers. Besides, Fig. 7 illustrates the error bound.
Remark 9
In the same time, compared with the literature [45], under the event-triggered mechanism, Table 1 shows that the event-triggered control proposed in this paper can better reduce the number of triggers.
Remark 10
Reference [6] studied the quasi-synchronization problem of a periodic time-varying complex dynamic network under the event-triggered control strategy. In contrast, the adaptive event-triggered control strategy proposed in this paper can reduce the triggering times more effectively and avoid unnecessary resource waste.
Example 2
As a typical nonlinear system, Chua’s circuit system is used to verify the reliability of the above conclusions. Let us consider a CDN with 5 nodes and 3 topologies. Here, the system parameters are set to:
\(A = \left[ {\begin{array}{*{20}{c}} { - 5}&{}1&{}0&{}{ - 1}&{}0&{}0&{}1&{}0&{}0&{}0&{}0&{}0&{}{ - 1}&{}{ - 1}&{}{ - 1}\\ 1&{}{ - 5}&{}0&{}0&{}{ - 1}&{}0&{}0&{}1&{}1&{}0&{}0&{}0&{}{ - 1}&{}0&{}0\\ 0&{}0&{}{ - 5}&{}0&{}0&{}0&{}0&{}0&{}0&{}1&{}0&{}0&{}{ - 1}&{}0&{}{ - 1}\\ { - 1}&{}0&{}0&{}{ - 5}&{}0&{}0&{}0&{}0&{}0&{}0&{}1&{}0&{}0&{}0&{}0\\ 0&{}{ - 1}&{}0&{}0&{}{ - 5}&{}0&{}0&{}0&{}0&{}0&{}1&{}0&{}0&{}0&{}0\\ 0&{}0&{}0&{}0&{}0&{}{ - 5}&{}0&{}0&{}0&{}0&{}0&{}0&{}0&{}0&{}0\\ 1&{}0&{}0&{}0&{}0&{}0&{}{ - 5}&{}0&{}0&{}1&{}0&{}0&{}{ - 1}&{}0&{}0\\ 0&{}1&{}0&{}0&{}0&{}0&{}0&{}{ - 5}&{}0&{}0&{}0&{}0&{}0&{}0&{}1\\ 0&{}1&{}0&{}0&{}0&{}0&{}0&{}0&{}{ - 5}&{}0&{}0&{}0&{}0&{}0&{}0\\ 0&{}0&{}1&{}0&{}0&{}0&{}1&{}0&{}0&{}{ - 5}&{}0&{}0&{}0&{}0&{}0\\ 0&{}0&{}0&{}1&{}1&{}0&{}0&{}0&{}0&{}0&{}{ - 5}&{}0&{}0&{}0&{}0\\ 0&{}0&{}0&{}0&{}0&{}0&{}0&{}0&{}0&{}0&{}0&{}{ - 5}&{}0&{}0&{}0\\ { - 1}&{}{ - 1}&{}{ - 1}&{}0&{}0&{}0&{}{ - 1}&{}0&{}0&{}0&{}0&{}0&{}{ - 5}&{}1&{}1\\ { - 1}&{}0&{}0&{}0&{}0&{}0&{}0&{}0&{}0&{}0&{}0&{}0&{}1&{}{ - 5}&{}0\\ { - 1}&{}0&{}{ - 1}&{}0&{}0&{}0&{}0&{}1&{}0&{}0&{}0&{}0&{}1&{}0&{}{ - 5} \end{array}} \right] \), \(\Gamma = \left[ {\begin{array}{*{20}{c}} 1&{}0&{}0\\ 0&{}1&{}0\\ 0&{}0&{}1 \end{array}} \right] \), \( S = \left[ {\begin{array}{*{20}{c}} { - 1}&{}1&{}0&{}0&{}0\\ 1&{}{ - 2}&{}1&{}0&{}0\\ 1&{}0&{}{ - 2}&{}1&{}0\\ 1&{}0&{}0&{}{ - 2}&{}1\\ 1&{}1&{}0&{}0&{}{ - 2} \end{array}} \right] \).
Furthermore, we choose the nonlinear function as:
where \(\chi ({x_{i1}}(t)) = - 0.68 - 0.295(|{x_{i1}}(t) + 1| - |{x_{i1}}(t) - 1|)\).
It is similar to the Example 1, given the conditions, we chose the other parameters of the system as: \(\alpha = 1\), \(\sigma = 0.5\), \(\lambda = 1\), \(l = 2\). Not only that, we set \({x_1}(0) = {\left[ {\begin{array}{*{20}{c}} 2&{ - 3.4}&{0} \end{array}} \right] ^{\textrm{T}}}\), \({x_2}(0) = {\left[ {\begin{array}{*{20}{c}} 0&4&{ - 2} \end{array}} \right] ^{\textrm{T}}}\), \({x_3}(0) = {\left[ {\begin{array}{*{20}{c}} { - 4.5}&{ - 2.5}&5 \end{array}} \right] ^{\textrm{T}}}\), \({x_4}(0) = {\left[ {\begin{array}{*{20}{c}} { - 1}&{3.5}&{ - 4} \end{array}} \right] ^{\textrm{T}}}\) and \({x_5}(0) = {\left[ {\begin{array}{*{20}{c}} {1}&{ - 0.5}&{1.5} \end{array}} \right] ^{\textrm{T}}}\) as the initial conditions.
The control gain matrices obtained by calculating the LMI in the MATLAB toolbox and some parameters therein are shown below: \(\gamma = 0.5\), and
In terms of the above parameters, the simulation is carried out. As illustrated in Figs. 8, 9, 10, 11, and 12, the trajectories of states as well as the trajectories of the error system are stable under ETC. When Fig. 13 depicts a system that is not stable without a controller. Besides, the adaptive law and adaptive strength are shown in Figs. 14 and 15 respectively. As well as Fig. 16 expresses the event intervals and Fig. 17 shows the error bound.
5 Conclusions
In this paper, the exponential synchronization problem with time delay uncertain CDNs has been discussed. However, because CDN is susceptible to interference from external environment, its system is affected. Therefore, we have proposed an adaptive dynamic event-triggered control strategy to better adapt to different environmental changes and system states and reduce possible instability problems. In addition, it can save the sampling times and reduce the information transmission frequency effectively. Then, Lyapunov function and inequality technique have been used to realize the synchronization of CDN under the control of ETC, and sufficient conditions to make certain that the exponential synchronization of CDN have been obtained. In addition, it has also been proved that the system can avoid Zeno behavior. In the end, two simulation examples have been used to prove the reliability of the theory. In the future research, we will compare with reference [46] to study the exponential congruence problem of complex dynamic networks with multi-coupling delays.
Remark 11
The limitation of adaptive event-triggered control strategy is that in some cases, the system cannot accurately judge when the event is triggered. This may be because the system does not have enough information to judge, or because the triggering conditions of the event are very complex and variable.
Remark 12
Although the control strategy proposed in this paper can effectively reduce the number of triggers, the trigger of the event is not monitored in real time, and the real-time occurrence of the event cannot be guaranteed. Therefore, the effect of adaptive control will be affected. In future work, in order to improve such limitations, more sensor data will be considered. By increasing the type and quantity of sensor data, more information can be provided to support the decision of event-triggered control.
Data availability
All data generated or analyzed during this study are included in this published article (and its supplementary information files).
References
Zou L, Wang Z, Hu J, Dong H. Partial-node-based state estimation for delayed complex networks under intermittent measurement outliers: a multiple-order-holder approach. IEEE Trans Neural Netw Learn Syst 1–15
Chanthorn P, Rajchakit G, Thipcha J, Emharuethai C, Sriraman R, Lim CP, Ramachandran R (2020) Robust stability of complex-valued stochastic neural networks with time-varying delays and parameter uncertainties. Mathematics 8:742
Arbi A, Tahri N (2022) Almost anti-periodic solution of inertial neural networks model on time scales. MATEC Web Conf 355:02006
Sheng L, Niu YC, Wang WL, Gao M, Geng YF, Zhou DH (2022) Estimation of toolface for dynamic point-the-bit rotary steerable systems via nonlinear polynomial filtering. IEEE Trans Ind Electron 69:7192–7201
Arbi A, Tahri N (2022) Stability analysis of inertial neural networks: a case of almost anti-periodic environment. Math Methods Appl Sci 45:10476–10490
Hu J, Jia CQ, Liu HJ, Yi XJ, Liu YR (2021) A survey on state estimation of complex dynamical networks. Int J Syst Sci 52:3351–3367
Rajchakit G, Pratap A, Raja R, Cao J, Alzabut J, Huang C (2019) Hybrid control scheme for projective lag synchronization of RiemannCLiouville sense fractional order memristive BAM neural networks with mixed delays. Mathematics 7:759
Tang Z, Park JH, Wang Y, Feng JW (2019) Distributed impulsive quasi-synchronization of Lur’e networks with proportional delay. IEEE Trans Cybern 49:3105–3115
Wu YB, Li HZ, Li WX (2021) Intermittent control strategy for synchronization analysis of time-varying complex dynamical networks. IEEE Trans Syst Man Cybern-Syst 51:3251–3262
Zhao YS, Li XD, Rao RF (2021) Synchronization of nonidentical complex dynamical networks with unknown disturbances via observer-based sliding mode control. Neurocomputing 454:441–447
He XL, Zhang HY (2022) Exponential synchronization of complex networks via feedback control and periodically intermittent noise. J Frankl Inst 359:3614–3630
Gunasekaran N, Ali MS, Arik S, Ghaffar H, Diab A (2022) Finite-time and sampled-data synchronization of complex dynamical networks subject to average dwell-time switching signal. Neural Netw 149:137–145
Baluni S, Yadav VK, Das S (2022) Quasi projective synchronization of time varying delayed complex valued Cohen-Grossberg neural networks. Inf Sci 612:231–240
Kumar A, Das S, Baluni S, Yadav VK, Lu J (2022) Global quasi-synchronisation of fuzzy cellular neural networks with time varying delay and interaction terms. Int J Syst Sci 53:2679–2693
Wang MZ, Li XD, Duan PY (2022) Event-triggered delayed impulsive control for nonlinear systems with application to complex neural networks. Neural Netw 150:213–221
Hu TT, He Z, Zhang XJ, Zhong SM, Shi KB, Zhang YY (2022) Adaptive fuzzy control for quasi-synchronization of uncertain complex dynamical networks with time-varying topology via event-triggered communication strategy. Inf Sci 582:704–724
Xu ZW, Shi P, Su HY, Wu ZG, Huang TW (2018) Global \({H_\infty }\) pinning synchronization of complex networks with sampled-data communications. IEEE Trans Neural Netw Learn Syst 29:1467–1476
Li X, Wu H, Cao J (2021) Synchronization in finite time for variable-order fractional complex dynamic networks with multi-weights and discontinuous nodes based on sliding mode control strategy. Neural Netw 139:335–347
Tabuada P (2007) Event-triggered real-time scheduling of stabilizing control tasks. IEEE Trans Autom Control 52:1680–1685
Jeong J, Lim Y, Parivallal A (2023) An asymmetric Lyapunov-Krasovskii functional approach for event-triggered consensus of multi-agent systems with deception attacks. Appl Math Comput 439
Vadivel R, Ali MS, Joo YH (2020) Event-triggered \({H_\infty }\) synchronization for switched discrete time delayed recurrent neural networks with actuator constraints and nonlinear perturbations. J Frankl Inst-Eng Appl Math 357:4079–4108
Xu Y, Sun J, Wu Z, Wang G (2022) Fully distributed adaptive event-triggered control of networked systems with actuator bias faults. IEEE Trans Cybern 52:10773–10784
Vadivel R, Hammachukiattikul P, Zhu Q, Gunasekaran N (2023) Event-triggered synchronization for stochastic delayed neural networks: Passivity and passification case. Asian J Control 25:2681–2698
Lv XX, Cao JD, Li XD, Abdel-Aty M, Al-Juboori UA (2021) Synchronization analysis for complex dynamical networks with coupling delay via event-triggered delayed impulsive control. IEEE Trans Cybern 51:5269–5278
Vadivel R, Hammachukiattikul P, Gunasekaran N, Saravanakumar R, Dutta H (2021) Strict dissipativity synchronization for delayed static neural networks: an event-triggered scheme. Chaos Solitons Fractals 150
Vadivel R, Joo YH (2020) Robust event-triggered T-S fuzzy system with successive time-delay signals and its application. IET Control Theory Appl 14:3697–3712
Niu Y, Sheng L, Gao M, Zhou D (2021) Dynamic event-triggered state estimation for continuous-time polynomial nonlinear systems with external disturbances. IEEE Trans Ind Inf 17:3962–3970
Bai J, Wu H, Cao J (2022) Topology identification for fractional complex networks with synchronization in finite time based on adaptive observers and event-triggered control. Neurocomputing 505:166–177
Wu ZG, Shi P, Su HY (2014) Sampled-data fuzzy control of chaotic systems based on a T-S Fuzzy model. IEEE Trans Fuzzy Syst 22:153–163
Ahmed-Ali T, Fridman E, Giri F (2016) Using exponential time-varying gains for sampled-data stabilization and estimation. Automatica 67:244–251
Gassara H, Hajjaji AE, Chaabane M (2010) Delay-dependent \({H_\infty }\) exponential stabilization of T-S fuzzy systems with interval time-varying delay. IEEE Decis Control 49:4281–4287
Park PG, Lee WI, Lee SY (2015) Auxiliary function-based integral inequalities for quadratic functions and their applications to time-delay systems. J Frankl Inst 352:1378–1396
Park MJ, Kwon OM, Ryu JH (2017) Advanced stability criteria for linear systems with time-varying delays. J Frankl Inst 355:520–543
Rajchakit G, Sriraman R (2021) Robust passivity and stability analysis of uncertain complex-valued impulsive neural networks with time-varying delays. Neural Process Lett 53:581–606
Chanthorn P, Rajchakit G, Humphries U, Kaewmesri P, Sriraman R, Lim CP (2020) A delay-dividing approach to robust stability of uncertain stochastic complex-valued hopfield delayed neural networks. Symmetry 12:683
Arbi A, Cao J, Alsaedi A (2018) Improved synchronization analysis of competitive neural networks with time-varying delays. Nonlinear Anal-Model Control 23:82–102
Arbi A, Cao J, Es-saiydy M, Zarhouni M, Zitane M (2022) Dynamics of delayed cellular neural networks in the Stepanov pseudo almost automorphic space. Discrete Contin Dyn Syst-S 15:3097–3109
Lien CH, Yu KW, Chen WD (2007) Stability criteria for uncertain Takagi-Sugeno fuzzy systems with interval time-varying delay. Control Theory Appl Iet 1:764–769
Li L, Liu XD, Chai TY (2009) New approaches on \({H_\infty }\) control of T-S fuzzy systems with interval time-varying delay. Fuzzy Sets Syst 160:1669–1688
Dong T, Wang A, Zhu H, Liao X (2018) Event-triggered synchronization for reaction-diffusion complex networks via random sampling. Physica A 495:454–462
Behinfaraz R, Ghaemi S, Khanmohammadi S, Badamchizadeh MA (2022) Time-varying parameters identification and synchronization of switching complex networks using the adaptive fuzzy-impulsive control with an application to secure communication. Asian J Control 24:377–387
Hu TT, Park JH, He Z, Zhang XJ, Zhong SM (2022) State-based event-triggered consensus strategy for Takagi-Sugeno fuzzy fractional-order multiagent systems with switching topologies. ISA Trans 126:109–120
Shi JY, Zhou PP, Cai SM, Jia Q (2023) Exponential synchronization for multi-weighted dynamic networks via finite-level quantized control with adaptive scaling gain. Chaos Solitons Fractals 174:113798
Peng D, Li X (2020) Leader-following synchronization of complex dynamic networks via event-triggered impulsive control. Neurocomputing 412:1–10
Bai J, Wu H, Cao J (2022) Topology identification for fractional complex networks with synchronization in finite time based on adaptive observers and event-triggered control. Neurocomputing 505:166–177
Baluni S, Das S, Yadav VK, Cao J (2022) Lagrange \(\alpha \)-exponential synchronization of non-identical fractional-order complex-valued neural networks. Circuits Syst Signal Process 41:5632–5652
Acknowledgements
This work is partially supported by the National Natural Science Foundation of China No. 61273004 and the Natural Science Foundation of Hebei Province No. F2021203061.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no known competing conflicts of interest in this paper. We declare that we do not have any economic conflicts or commercial interests related to the submitted works.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This work is supported by National Natural Science Foundation of China (No. 61273004) and the Natural Science Foundation of Hebei province (No. F2021203061).
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Zhao, Y., Ma, Y. Exponential Synchronization of Complex Dynamic Networks with Time Delay and Uncertainty via Adaptive Event-Triggered Control. Neural Process Lett 56, 81 (2024). https://doi.org/10.1007/s11063-024-11550-y
Accepted:
Published:
DOI: https://doi.org/10.1007/s11063-024-11550-y