Zhang Wenguang, Yan Jin
School of Information Engineering, Xuzhou University of Technology, Lishui Road 2, Xuzhou, China.
School of Computer Science and Engineering, Nanjing University of Science and Technology, Xiaolingwei 200, Nanjing, China.
ISA Trans. 2024 Feb;145:148-162. doi: 10.1016/j.isatra.2023.11.030. Epub 2023 Nov 20.
In this paper, an innovative event-triggered optimal tracking control algorithm is proposed for input saturated strict-feedback nonlinear systems with unknown dynamics. In order to reduce the requirement of configuring a complete suit of sensors and enhance the reliability of the controlled system, a neural networks (NNs) based adaptive state observer is developed firstly to reconstruct the system states. Subsequently, based on the state estimation information, a hybrid-triggered feedforward controller is designed to transform the original tracking control problem into an equivalent regulation issue, which is then solved by developing an event-triggered optimal controller. Therefore, the final controller consists of a hybrid-triggered feedforward controller and an event-triggered optimal controller. In order to make the actual input signals of the two controllers be updated simultaneously, a synchronization-oriented triggering rule is established by using multiple triggering errors. By virtue of this unique framework, the proposed control scheme can not only minimize the predefined cost function, but also greatly reduce the data transmission. What is more, the convergence properties of the proposed control strategy are achieved by using Lyapunov theory. It is important to note that unlike the widely adopted observer-controller framework, where the separation principle holds for the design of the state observer, there is a considerable coupling relationship between the error dynamics of the state observer and the event-triggered optimal controller in this paper. The distinguishing feature of the proposed method is its ability to ensure a satisfactory level of precision in both state estimation and tracking control, even in the presence of control saturation issues. At last, the proposed control strategy is applied to the tracking control problem of a high-order robot system and marine surface vehicle to demonstrate its effectiveness.
本文针对具有未知动态特性的输入饱和严格反馈非线性系统，提出了一种创新的事件触发最优跟踪控制算法。为了降低配置一整套传感器的要求并提高受控系统的可靠性，首先开发了一种基于神经网络（NNs）的自适应状态观测器来重构系统状态。随后，基于状态估计信息，设计了一种混合触发前馈控制器，将原跟踪控制问题转化为一个等效调节问题，然后通过开发一个事件触发最优控制器来解决该问题。因此，最终的控制器由混合触发前馈控制器和事件触发最优控制器组成。为了使两个控制器的实际输入信号同时更新，利用多个触发误差建立了一种面向同步的触发规则。借助这一独特框架，所提出的控制方案不仅可以使预定义的代价函数最小化，还能大大减少数据传输。此外，利用李雅普诺夫理论实现了所提出控制策略的收敛特性。需要注意的是，与广泛采用的观测器 - 控制器框架不同，在该框架中状态观测器的设计遵循分离原理，而本文中状态观测器的误差动态与事件触发最优控制器之间存在相当大的耦合关系。所提方法的显著特点是即使在存在控制饱和问题的情况下，也能在状态估计和跟踪控制方面确保令人满意的精度水平。最后，将所提出的控制策略应用于高阶机器人系统和水面舰艇的跟踪控制问题，以证明其有效性。
