A Prescribed Performance Fully-Actuated Control Approach Based on Adaptive Neural Network

This paper addresses the tracking control problem of a second-order system by designing a novel adaptive neural network-based prescribed performance control (PPC) approach within the framework of a fully-actuated system (FAS). The PPC method is utilized to ensure that tracking errors are constrained within predefined bounds, after which the system is linearized using the FAS approach. Besides, the unknown external disturbance is approximated using a radial basis function neural network (RBF NN). The adaptive neural network can online update the estimation of the disturbance to maintain the closed-loop system's convergence through the system state, thereby guaranteeing the effectiveness of the control approach via the Lyapunov stability theory. The advantage of our proposed approach over the existing exponential performance function is illustrated through numerical simulations on a two-link manipulator system.