This paper deals with the parameter identification problem of Hammerstein output error auto-regressive (OEAR) systems with different nonlinearities by combining the key-term separation principle and the auxiliary model identification idea. The basic idea is, by using the key-term separation principl