基于神经网络的非线性随机系统有限时间环域稳定性与镇定  

Finite-Time Annular Domain Stability and Stabilization for Nonlinear Stochastic Systems Via Neural Network

作　　者：严志国 苏方旭 陈耀 张敏[1] YAN Zhi-guo;SU Fang-xu;CHEN Yao;ZHANG Min(School of Electrical Engineering and Automation,Qilu University of Technology(Shandong Academy of Sciences),Jinan 250353,China;School of Mathematics and Statistics,Qilu University of Technology(Shandong Academy of Sciences),Jinan 250353,China)

机构地区：[1]齐鲁工业大学(山东省科学院)电气工程与自动化学院,济南250353 [2]齐鲁工业大学(山东省科学院)数学与统计学院,济南250353

出　　处：《齐鲁工业大学学报》2021年第4期32-42,共11页Journal of Qilu University of Technology

基　　金：国家自然科学基金(61877062,61977043)。

摘　　要：研究了一类不确定非线性随机系统的有限时间环域稳定性与镇定问题。首先,给出了不确定非线性随机系统有限时间环域稳定性的定义。其次,根据多层反馈神经网络对非线性项进行逼近,并得到了该系统为有限时间环域稳定的一个充分条件。再次,设计了状态反馈有限时间环域控制器,得到了一个由矩阵不等式所表示的充分条件,并给出了相应的求解算法。最后,通过一个算例说明了上述结果的有效性。In this paper,the problem of the finite-time annular domain stability and stabilization for uncertain nonlinear stochastic systems is investigated.First,the definition of finite-time annular domain stability of uncertain nonlinear stochastic systems is given.Second,the nonlinearities are approximated by using multi-layer feedback neural networks technique,and a sufficient condition for testing finite-time annular domain stability of a system is obtained.Third,a finite-time annular domain state feedback controller is derived by using matrix inequality method and an algorithm for solving the corresponding to matrix inequalities is given.Finally,an example is given to illustrate the effectiveness of the above results.

关 键 词：随机系统 有限时间环域稳定性 神经网络 

分 类 号：TP273.5[自动化与计算机技术—检测技术与自动化装置]