Composite Adaptive Control of Nonlinear Systems with Periodic Disturbances

For dealing with the periodic disturbances, the FNNs are frequently incorporated with the Fourier neural networks (FNNs) in the existing schemes for compensating the periodic disturbances and the unknown nonlinearity simultaneously. However, the periods of the disturbance signals must be known a priori in these works. To relax such a restriction, a composite adaptive FNN control for linearizable systems with periodic disturbances is synthesized in this paper. By a mild assumptions on the unknown disturbances, the proposed design ensures uniform convergence of the tracking error and the prediction errors to the vicinity of zero, whose size can be made arbitrarily small via using sufficiently large gains. Simulation studies demonstrating its validity are carried out in the final.