Block control design based on multiplayer neural network

For a class of linear systems with mismatched uncertainties of unknown bound, a robust control method which combines block control, neural network control and backstepping techniques is proposed based on the block control principle. The bounds of the uncertainties are not required to be known. The mismatched uncertainties are overcome by using the backstepping technique. The uncertainties are estimated by multiplayer neural network (MNN) approximators. The performance of the system is improved by using robust control. The stability of the closed-loop system is proved in the sense of Lyapunov stability theorem. Simulation examples have shown the rightness and effectiveness of the proposed scheme.