Sliding polynomial modeling has undergone considerable development and has been widely applied to dynamic modeling. One focus of the polynomial model is to make efficient use of prior information, which can also be called additional or constrained information. Considering the fact that much additional information can be translated to equality constraints, we develop a new two-stage approach for sliding polynomial modeling with equality constraints. To maintain its compatibility with the polyno-mial model without this constraint, only a simple adjustment based on the constraint equation is involved in the algorithm. The approach is easy to implement and is superior in observability, convergence and accuracy. The results of a simulation study of spacecraft tracking are provided to confirm the theoretical development. As no special hypotheses are required, the approach presented here can be widely applied to dynamic modeling for nonlinear systems with equality constraints. Furthermore, per-formance predictions for other dynamic models can also draw lessons from this approach.
