The event-triggered consensus control problem is studied for nonstrict-feedback nonlinear systems with a dynamic leader. Neural networks (NNs) are utilized to approximate the unknown dynamics of each follower and its neighbors. A novel adaptive event-trigger condition is constructed, which depends on the relative output measurement, the NN weights estimations, and the states of each follower. Based on the designed event-trigger condition, an adaptive NN controller is developed by using the backstepping control design technique. In the control design process, the algebraic loop problem is overcome by utilizing the property of NN basis functions and by designing novel adaptive parameter laws of the NN weights. The proposed adaptive NN event-triggered controller does not need continuous communication among neighboring agents, and it can substantially reduce the data communication and the frequency of the controller updates. It is proven that ultimately bounded leader-following consensus is achieved without exhibiting the Zeno behavior. The effectiveness of the theoretical results is verified through simulation studies.

中文翻译：

---

非严格反馈非线性系统的基于神经网络的自适应事件触发共识控制。

对于具有动态前导的非严格反馈非线性系统，研究了事件触发的共识控制问题。利用神经网络（NN）来估算每个跟随者及其邻居的未知动态。构建了一种新颖的自适应事件触发条件，该条件取决于相对输出测量，NN权重估计以及每个跟随者的状态。基于设计的事件触发条件，利用反步控制设计技术开发了自适应神经网络控制器。在控制设计过程中，通过利用NN基函数的特性并设计新的NN权重自适应参数定律来克服代数环问题。所提出的自适应NN事件触发控制器不需要相邻代理之间的连续通信，并且可以大大减少数据通信和控制器更新的频率。事实证明，在不表现出芝诺行为的情况下，最终达成了领导者以下的共识。通过仿真研究验证了理论结果的有效性。