Abstract: The foundation of Steady-State models for optimizing control of steady-state industrial processes which are described by bilinear models uses the step signals of steady-state as input identification signals in the course of optimizing control and obtain the strong consistency estimations of steady-state models. But only under the hypothesis of the Fi is reversible matrix, the identification technique is used. In the paper, the reversibility of Fi is studied and the sufficient conditions for system identifiability are given.