This article addresses the adaptive neural tracking control problem for a class of uncertain stochastic nonlinear systems with nonstrict-feedback form and prespecified tracking accuracy. Some radial basis function neural networks (RBF NNs) are used to approximate the unknown continuous functions online, and the desired controller is designed via the adaptive dynamic surface control (DSC) method and the gain suppressing inequality technique. Different from the reported works on uncertain stochastic systems, by combining some non-negative switching functions and dynamic surface method with the nonlinear filter, the design difficulty is overcome, and the control performance is analyzed by employing stochastic Barbalat’s lemma. Under the constructed controller, the tracking error converges to the accuracy defined a priori in probability. The simulation results are shown to verify the availability of the presented control scheme.

中文翻译：

---

不确定随机非严格反馈系统具有预定跟踪精度的自适应神经动态表面控制

本文解决了一类具有非严格反馈形式和预定跟踪精度的不确定随机非线性系统的自适应神经跟踪控制问题。一些径向基函数神经网络（RBF NNs）用于在线逼近未知连续函数，并通过自适应动态表面控制（DSC）方法和增益抑制不等式技术设计所需的控制器。与已报道的关于不确定随机系统的工作不同，通过将一些非负切换函数和动态曲面法与非线性滤波器相结合，克服了设计困难，并采用随机Barbalat引理分析了控制性能。在构建的控制器下，跟踪误差收敛到定义的精度先验概率。显示仿真结果以验证所提出的控制方案的可用性。