Adaptive neural control of MIMO uncertain nonlinear systems with unmodeled dynamics and output constraint

In this paper, an adaptive dynamic surface control (DSC) method is investigated for a class of multiple-input-multiple-output block-structure uncertain nonlinear systems with unmodeled dynamics and output constraint. Modified DSC is used to construct the vector virtual control and control strategy in recursive each step. A dynamic signal is used to handle unmodeled dynamics in the system. Radial basis function neural networks are used to estimate the unknown continuous black-box functions. By introducing the symmetric barrier Lyapunov functions, the constraint requirement of each element in the output is guaranteed in the designed adaptive control system. Using the defined compact set in the theoretical analysis and the proof characteristics of DSC, the introduced input uncertain term is effectively deal with in the controller design. The stability analysis shows that all the signals in the closed-loop control system are semiglobally uniformly ultimately bounded. A numerical simulation example with the two-link flexible robotic system is provided to verify the effectiveness of the constructed control scheme.