Ding Sanbo, Wang Zhanshan, Xie Xiangpeng
IEEE Trans Neural Netw Learn Syst. 2022 Aug;33(8):3622-3633. doi: 10.1109/TNNLS.2021.3053652. Epub 2022 Aug 3.
In this article, we investigate the periodic event-triggered synchronization of discrete-time complex dynamical networks (CDNs). First, a discrete-time version of periodic event-triggered mechanism (ETM) is proposed, under which the sensors sample the signals in a periodic manner. But whether the sampling signals are transmitted to controllers or not is determined by a predefined periodic ETM. Compared with the common ETMs in the field of discrete-time systems, the proposed method avoids monitoring the measurements point-to-point and enlarges the lower bound of the inter-event intervals. As a result, it is beneficial to save both the energy and communication resources. Second, the "discontinuous" Lyapunov functionals are constructed to deal with the sawtooth constraint of sampling signals. The functionals can be viewed as the discrete-time extension for those discontinuous ones in continuous-time fields. Third, sufficient conditions for the ultimately bounded synchronization are derived for the discrete-time CDNs with or without considering communication delays, respectively. A calculation method for simultaneously designing the triggering parameter and control gains is developed such that the estimation of error level is accurate as much as possible. Finally, the simulation examples are presented to show the effectiveness and improvements of the proposed method.
在本文中，我们研究离散时间复杂动态网络（CDNs）的周期性事件触发同步。首先，提出了一种离散时间版本的周期性事件触发机制（ETM），在该机制下传感器以周期性方式对信号进行采样。但是采样信号是否传输到控制器由预定义的周期性ETM决定。与离散时间系统领域中的常见ETM相比，所提出的方法避免了逐点监测测量值，并扩大了事件间间隔的下限。因此，有利于节省能量和通信资源。其次，构造“不连续”的李雅普诺夫泛函来处理采样信号的锯齿状约束。这些泛函可被视为连续时间领域中那些不连续泛函的离散时间扩展。第三，分别针对考虑或不考虑通信延迟的离散时间CDNs，推导了最终有界同步的充分条件。开发了一种同时设计触发参数和控制增益的计算方法，以使误差水平的估计尽可能准确。最后，给出仿真示例以展示所提方法的有效性和改进之处。
