Song Zhibao, Gao Lihong, Wang Zhen, Li Ping
IEEE Trans Neural Netw Learn Syst. 2024 Dec;35(12):18771-18783. doi: 10.1109/TNNLS.2023.3321596. Epub 2024 Dec 2.
In this article, the adaptive neural control is studied for multiple-input-multiple-output (MIMO) nonlinear systems with asymmetric input saturation, dead zone, and full state-function constraints. A suitable transformation is introduced to overcome the dead zone and saturation nonlinearity, and radial basis function (RBF) neural networks (NNs) are used to approximate the unknown nonlinear functions. What is more, we apply the Nussbaum function and time-varying barrier Lyapunov function (BLF) to deal with the unknown control gains and full state-function constraints, respectively. Based on the backstepping method, a universal adaptive neural control scheme is presented such that not only the state-function constraints of the closed-loop system cannot be violated and all signals of the closed-loop systems are bounded, but also the tracking error converges to a small neighborhood containing the origin. The effectiveness of the proposed control scheme is verified by an application to the mass-spring-damper system and a numerical example.
本文研究了具有非对称输入饱和、死区和全状态函数约束的多输入多输出（MIMO）非线性系统的自适应神经控制。引入了一种合适的变换来克服死区和饱和非线性，并用径向基函数（RBF）神经网络（NNs）逼近未知非线性函数。此外，我们分别应用努斯鲍姆函数和时变障碍李雅普诺夫函数（BLF）来处理未知控制增益和全状态函数约束。基于反步法，提出了一种通用的自适应神经控制方案，使得闭环系统不仅不会违反状态函数约束且闭环系统的所有信号都是有界的，而且跟踪误差收敛到包含原点的一个小邻域内。通过应用于质量 - 弹簧 - 阻尼器系统和一个数值例子验证了所提出控制方案的有效性。
