Parameter identification of continuous - time Hammerstein system from step responses

Direct identification is studied for the continuous-time Hammerstein system. The step signal is used as the excitation input, its responses and the white noises are superimposed as the system output. First, the regression equation is derived from the multiple integrals of the system inputs and outputs, whose parameters are estimated by least square algorithm and instrument variable (IV) technique. Then, the static coefficient LSE (least-squares estimation) is solved from the regression equation from the amplitudes of the signal, system stable outputs and nonlinear coefficients. The effectiveness of the method is proved through three examples.