Gao Tingting, Liu Yan-Jun, Li Dapeng, Tong Shaocheng, Li Tieshan
IEEE Trans Cybern. 2021 Apr;51(4):1943-1953. doi: 10.1109/TCYB.2019.2906118. Epub 2021 Mar 17.
In this paper, an adaptive neural network (NN) control scheme is developed for a class of stochastic nonlinear systems with time-varying full state constraints. In the controller design, RBF NNs are employed to approximate the unknown terms, and the backtracking technique is introduced to overcome the restriction of matching conditions. At the same time, tangent type time-varying barrier Lyapunov functions (tan-TVBLFs) are constructed to ensure the full state constraints are never violated, where tan-TVBLFs are beneficial to integrate constraint analysis into a common method. Furthermore, the Lyapunov stability theory is used to prove that all closed-loop signals are semiglobal uniformly ultimately bounded in probability and error signals remain in the compact set do not violate the time-varying constraints. A simulation example will be used to exhibit the effectiveness of the proposed control scheme.
本文针对一类具有时变全状态约束的随机非线性系统，提出了一种自适应神经网络（NN）控制方案。在控制器设计中，采用径向基函数神经网络（RBF NNs）逼近未知项，并引入回溯技术克服匹配条件的限制。同时，构造了切线型时变障碍李雅普诺夫函数（tan-TVBLFs）以确保全状态约束永远不会被违反，其中tan-TVBLFs有利于将约束分析集成到一种通用方法中。此外，利用李雅普诺夫稳定性理论证明了所有闭环信号在概率上是半全局一致最终有界的，并且误差信号保持在不违反时变约束的紧致集中。将通过一个仿真例子来展示所提出控制方案的有效性。
