Abstract: Although the existing balance filtering algorithm for calculating the high altitude airship gesture angle has fused the acceleration information and the angular velocity information， it did not analyze the application conditions of the acceleration， resulting in accumulation errors and dynamic solver accuracy problems. In order to solve these problems， a selfadaption gesture calculating algorithm was put forward. According to the airship’s different motion states， the proposed algorithm integrated the acceleration and angular velocity data to adjust the gesture calculate algorithm. The validity of this algorithm was verified by the Matlab experiments， and the experimental results showed that this algorithm had effectively solved the weakness of the existing balance filtering algorithm. As introducing the concept of SMV (signal magnitude vector)， the high altitude airships’ motion states were judged accurately， and it also improved the accumulation errors problem and enhanced the dynamic solver accuracy of the high altitude airship.

Key words: information fusion, strapdown inertial navigation system (SINS), gesture calculating, high altitude airship, selfadaption