Robust observer backstepping neural network control of nonlinear systems in strict feedback form

An observer for nonlinear systems in strict feedback form is presented. Together with a backstepping neural network controller, the resulting control strategy can be used in systems where only the output is measurable and unmatched additive disturbances with unknown bounds exist. Some assumptions about the system are required in order to apply the separation principle. A stability proof is provided together with the simulation results.