For a class of nonlinear discrete-time dynamic systems, the intelligent modeling method of the nonlinear system is studied. Firstly, the forgetting factor recursive least square method is used to identify the unknown parameters of the lower-order model, and then the stochastic configuration networks are used to estimate the higher-order nonlinear part. Finally, the two identification methods are used to identify the system alternately under the external error criterion. Furthermore, an improved intelligent modeling method for alternating identification of nonlinear systems is proposed. The combination of the stochastic configuration networks and the recursive least square algorithm can effectively improve the identification accuracy of nonlinear systems, and the effectiveness of the proposed algorithm is verified by comparative analysis of numerical simulation experiments.