In the practical application of the BD—2 /MIMU tightly integrated navigation system

parameter of the inertial components is uncertain and BD—2 satellite navigation system vulnerable to interference

so the traditional Kalman filtering method can’t meet the requirements.A new robust adaptive filtering algorithm which is the innovation-based adaptive interval Kalman filtering（IAIKF） is put forward.In this algorithm

the model of the BD— 2 / MIMU tightly integrated navigation system is described as interval uncertain system

and gain is adjusted according to online information

to effectively solve the measurement error estimation variance when dramatic changes rapidly increases and system parameter uncertainty of filtered estimation influence.The simulation result demonstrates that this method is feasible

and because the algorithm provides bounds of the navigation results

it is very important to design integrated navigation system.