Abstract This paper addresses a novel finite-time adaptive neural control (FTANC) problem for nonstrict-feedback stochastic nonlinear systems (NSFSNS), in which the input delay and output constrained problems are considered simultaneously. First, the Pade approximation technique is adopted to transform the delay input system into a delay-free one. Second, the stochastic nonlinear mapping technique is developed to solve the symmetric and asymmetric output constraints in the system. Then, the adaptive neural controller is designed based on backstepping technique, such that the closed-loop systems are semi-globally practical finite-time stable (SGPFTS) in probability, and the tracking error converges to a small neighborhood of the origin after a finite period of time. Two simulation examples show the effectiveness of the proposed approach.

中文翻译：

---

具有输入延迟和输出约束的非严格反馈随机非线性系统的有限时间自适应神经控制

摘要 本文针对非严格反馈随机非线性系统 (NSFSNS) 提出了一种新颖的有限时间自适应神经控制 (FTANC) 问题，其中同时考虑了输入延迟和输出约束问题。首先，采用Pade逼近技术将延迟输入系统转化为无延迟系统。其次，开发了随机非线性映射技术来解决系统中的对称和非对称输出约束。然后，基于backstepping技术设计了自适应神经控制器，使得闭环系统在概率上是半全局实用有限时间稳定（SGPFTS），并且跟踪误差在有限时间后收敛到原点的一个小邻域。一段的时间。两个仿真实例显示了所提出方法的有效性。