The unscented incremental filter(UIF)model and analysis method were put forward

in which its concept

basis equations and the recursive calculative steps were established

and special concern was given to UIF with additive noise. In practice

the measurement data have unknown system errors

for the environmental factors

the instability of measurement equipments and the improper models and parameters. Under these conditions

the classic unscented Kalman filter(UKF)have greater filtering error

and even lead to diverge. The presented UIF can successfully eliminate these unknown system errors and improve the filtering accuracy. The method is simple to calculate and easy to apply in engineering.