Control of nonlinear systems with predefined constraints using neural networks

This paper proposes a new nonlinear mapping to address the output constraint problem. We transform the constrained tracking error into an equivalent unconstrained one. Then adaptive neural network (NN) control with predefined constraints is studied for nonlinear systems. The proposed scheme guarantees that all the signals in the closed-loop system are bounded and the system output asymptotically tracks the reference trajectory without the violation of the predefined constraints. Finally, we give a numerical example to show effectiveness of the proposed scheme.