An identification method based on combined signal sources is proposed to identify the nonlinear Hammerstein-Wiener model with coloured noise. The combined signal sources composed of separable signal and random signal are used to realize the separation of the parameter identification of the series modules for the Hammerstein-Wiener model with coloured noise, which can effectively simplifies the identification process. Firstly, based on the input and output of separable signal, the correlation analysis method is employed to suppress the interference of process noise and identify the parameters of the static nonlinear module and the dynamic linear module. Secondly, by means of the auxiliary model technique, the output of the auxiliary model and residual estimation are used to replace the unmeasurable intermediate variables and noise variables in the identification model, respectively. The recursive extended least square method based on auxiliary model is deduced to identify the parameters of input static nonlinear module and noise model according to the input and output data of random signal. Theoretical analysis and simulation results show that the proposed method can effectively identify the nonlinear Hammerstein-Wiener model with coloured noise and has good robustness.