Design of robust sliding mode observer based on RBF neural network

A robust sliding mode observer for the nonlinearities or uncertainties of systems is proposed. The sliding mode manifold and control methodology are proposed. The design of the observer's parameters needs not to solve a lot of equations. The proposed observer is robust to the nonlinearities or uncertainties of systems. An adaptive RBF neural network is then used to learn the upper bound of system uncertainties. The convergence rate between the observer and the system can be changed by choosing suitable sliding mode manifold, so as to attain the desired performances. Simulation results are presented to validate the design.