The adaptive filtering algorithms based on the polynomial model are widely used in the field of maneuvering target tracking, but there is no uniform evaluation criterion to measure the quality of these tracking algorithms. Due to the existence of time-varying unknown inputs, the maneuvering target state estimation is actually biased. To solve this problem, the minimum mean square error bound calculation method for polynomial model Kalman filters was derived based on the minimum mean square error criterion, and the process noise variance law minimizing the state estimation mean square error was obtained. The proposed algorithm provides a unified evaluation standard for maneuvering target tracking algorithms based on the polynomial model, and also provides the basis for the setting of the actual process noise variance in maneuvering target tracking. The effectiveness of the proposed algorithm is demonstrated by the simulation results.