A new online identification method is presented. The identified nonlinear s
ystems have partial-state measurement. Their inner states, parameters and s
tructures are unknown. The design is based on the combination of a model-fr
ee state observer and a neuro identifier. First, a sliding mode observer, w
hich does not need any information about the nonlinear system, is applied t
o obtain the full states. A dynamic multilayer neural network is then used
to identify the whole nonlinear system. The main contributions of the paper
are: a new observer-based identification algorithm is proposed; and a stab
le learning algorithm for the neuro identifier is given.