Neural Adaptive Optimal Control of Inequality-Constrained Nonlinear System With Partial Uncertain Time Delay
针对带有部分不确定时滞和不等式约束的离散时间非线性系统，提出了一种神经自适应最优跟踪控制方法，通过变换约束和时滞信息，利用神经网络逼近最优控制器和成本函数，并保证闭环信号有界。
An optimal tracking control system using neural adaptive techniques is introduced for nonlinear systems subjected to time delay and inequality constraints, which is partially uncertain. The nonlinear inequality constraints and partial uncertain time delay of the state are considered in the discrete-time nonlinear system. By transforming the inequality constraint information into augmented system state variables, and using the precompensator method, an augmentation system that contains constraints and transformed controller information is obtained. The Lyapunov–Krasovskii functionals (LKFs) can be used to deal with the partial uncertain state time delay. Subsequently, the optimal controller, the long-term cost function, the uncertain resistance, and system dynamics can be approximated by the action, critic, the disturbance, and the state estimation NNs, and suitable adaptive laws are obtained. Furthermore, the uniform ultimate boundedness (UUB) of the signals in the closed-loop control system can be obtained by the designed near-optimal controller. The inequality constraints are satisfied and the challenge arising from partial uncertain time delay has been successfully addressed, while a numerical simulation verification example is presented.