A new double-adaptive federated filter is put forward in order to get rid of the disadvantage of general federated Kalman filter that the statistic characteristics of the noise must be known exactly.Considering in multi-information integrated navigation in which the systematic noise of inertial navigation system is not known exactly and the measurement noise of satellite-navigation is time-varying

the new algorithm does not need to know the statistic characteristics of systematic noise and can adjust the measurement noise on-line

meanwhile

the information distribution coefficient can be adaptive adjusted according to the geometry dilution of precision（GDOP）

which is the real-time output of the satellite-navigation.In SINS-GPS-Galileo-BDS integrated navigation system

the simulation results which compared with general federated filter show that this method effectively improves the integrated navigation’s precision and reliability

and more applicable for the multi-information integrated navigation system which systematic noise is unknown and measurement noise is time-varying.