Dynamic Adaptive Attitude Estimation Method Based on Inertial Sensor*

Attitude estimation is of great significance for autonomous operation of unmanned systems. Inertial sensor is the main sensor for attitude measurement. It solves the attitude by gyro integration, and corrects the attitude by accelerometer. However, in the dynamic environment, the accelerometer is affected by the acceleration of the carrier motion, and its attitude correction will produce large errors. In view of this, this paper proposes a dynamic adaptive attitude estimation method based on the complementary filter architecture. It can evaluate the acceleration level of motion in real time and adjust the accelerometer correction coefficient in turn. At the same time, a gyro bias estimation method based on sliding window is designed, which can realize real-time gyro bias estimation under maneuvering conditions. Finally, this paper develops a set of dynamic attitude measurement system using the above methods. The experimental results demonstrate the effectiveness of the algorithm proposed in this paper in significantly enhancing the accuracy of attitude estimation of the carrier in maneuvering state.