Wang Jianhui, Liu Zhi, Zhang Yun, Chen C L Philip
IEEE Trans Neural Netw Learn Syst. 2019 Nov;30(11):3300-3312. doi: 10.1109/TNNLS.2018.2890699. Epub 2019 Jan 25.
In this paper, the uncertain direct of the hysteretic system component will be considered. Besides, the effect of stochastic disturbance inevitably exists in many practical systems, which would cause the instability. Simultaneously, it is significant to guarantee the perfect error tracking performance for the uncertain nonlinear hysteresis systems when operation suffers the failure. To ensure the maintaining acceptable system performance in reality, the new properties of the Nussbaum function are proposed, and an auxiliary virtual controller is designed through the neural network (NN) universal approximator. Furthermore, it is challenged to save the system-limited transmutation resource for nonlinear systems, especially for stochastic nonlinear systems, with unknown hysteresis input and actuator failures. The coupling effect of the system communication resource constrains has to arise the issue of the mutual coupling function, which makes that the tracking control design is more complicated. Using the proposed event-triggered controller and back-stepping technology, a new optimization algorithm is proposed to ensure that the states of the closed-loop system and the tracking error remain bounded in probability. Finally, to illustrate the effectiveness of our proposed adaptive NN control method with the event-triggered strategy, some numerical examples are provided.
本文将考虑滞后系统组件的不确定方向。此外，随机干扰的影响在许多实际系统中不可避免地存在，这会导致系统不稳定。同时，对于不确定非线性滞后系统，在运行出现故障时保证完美的误差跟踪性能具有重要意义。为确保在实际中维持可接受的系统性能，提出了努斯鲍姆函数的新特性，并通过神经网络（NN）通用逼近器设计了辅助虚拟控制器。此外，对于具有未知滞后输入和执行器故障的非线性系统，特别是随机非线性系统，节省系统有限的变换资源是一项挑战。系统通信资源约束的耦合效应必然引发相互耦合函数的问题，这使得跟踪控制设计更加复杂。利用所提出的事件触发控制器和反步技术，提出了一种新的优化算法，以确保闭环系统的状态和跟踪误差在概率上保持有界。最后，为了说明我们提出的具有事件触发策略的自适应神经网络控制方法的有效性，提供了一些数值例子。
