Projectiles Attitude Estimation Based On Multi-sensor Data Fusion

The measurement of the projectile’s attitudes plays a vital role in the development of precision-guided munitions. In this paper, the attitude angle of guided projectile is measured by combination three-axis magnetoresistive sensor with three-axis MEMS gyro. Study on the multi-sensor data fusion technology applications in magnetic-MEMS Gyro measurement attitude. The data fusion model based on the weighted method and kalman filtering method was established. Finally, data fusion method was proved feasible through numerical simulation and experiment.To analyze the merits and flaws of the existing measurement of projectile’s attitudes. The attitude angle of guided projectile measured by combination three-axis magnetoresistive sensor with three-axis MEMS gyro was proposed. The mathematical model of the attitude angle based on the magnetic-MEMS Gyro was established. The mathematical model of the initial attitude angle of the missile body was established, with not considering the yaw angle. The data fusion method of calculating projectile attitude angle based on weighted method was established and the weights are estimated by minimum variance.The state equation and measurement equation of the kalman filter was established by analyzing the sensor’s errors. Taking advantage of kalman filtering method to estimate the attitude angle error, it can feedback to correct gyro errors real-time and improve the projectile’s attitude angle precision. The proposed data fusion methods was verified by the simulation analysis and experimental verification. The simulation and experimental results show that weighted fusion method can improve the accuracy of attitude angle in a certain extent,and it can solve problem which magnetic resistance sensor can’t calculate attitude angle in the magnetic detection blind area. The data fusion method based on kalman filtering can estimate projectile flight attitude angle at a high precision. The data fusion method based on kalman filtering not only can get high precision attitude angle, but also make the attitude angle error not accumulate over time. So the data fusion method based on kalman filtering is much more superior than the data fusion method based on weight coefficient.