Predefined-Time and Predefined-Accuracy Adaptive Neural Tracking Control of Disturbed Nonstrict-Feedback Nonlinear Systems Under Asymmetric Output Constraints

This work concentrates on the predefined-time and predefined-accuracy tracking controller design problem for a nonstrict-feedback nonlinear system under asymmetric output constraints. Unlike finite-time or fixed-time control strategies in related literature, the great advantage of our presented controller design algorithm is that the algorithm can specify the convergence time and the accuracy of the system in advance, depending on the user’s requirements. First, an extremely important performance of RBF NNs is employed to handle the design obstacle caused by nonstrict-feedback structure. Then, a time-varying nonlinear transformation function is introduced to handle the asymmetric output constraints problems. Second, under the framework of adaptive backstepping design, we propose the predefined-time controller and the asymptotic tracking controller simultaneously, and a continuous switching function is used to transition the predefined-time controller to the asymptotic tracking controller when the resulting tracking error arrives at the specified accuracy. Under our control scheme, it is rigorously demonstrated that all signals of the studied system remain bounded and the tracking error achieves a predefined accuracy within a preset time. Finally, two practical simulation experiments are given to validate the effectiveness of the controller.