中文摘要：

构建了基于MEMS技术的陀螺、加速度计、磁强计及空速计组合的微小型飞行器姿态测量系统。研究了基于四元数的扩展卡尔曼滤波算法。取姿态误差四元数和陀螺随机漂移构建状态向量,通过误差四元数微分方程和陀螺随机误差模型建立卡尔曼滤波状态方程,采用速度信息实时补偿加速度计输出值得到重力矢量,利用重力矢量估计水平姿态,通过滤波补偿姿态误差,降低了对陀螺的精度要求。将状态向量之间的约束方程作为伪量测方程引入到量测模型中,解决了由于状态向量相互约束导致的滤波发散和奇异。动态飞行滤波噪声的自适应调整增强了系统性能。仿真和实验表明,该滤波算法能够有效避免系统的漂移,提高系统测量精度和稳定性。

英文摘要：

The attitude measurement system was constructed for micro aerial vehicles （MAV） based on the MEMS gyroscope, accelerometer, magnetometer and airspeed meter. The quatemion-based extended Kalman filter was investigated. Attitude quaternion errors and drift bias of gyroscope were selected to construct state vector, and then the state equation was established through quaternion error differential equation and stochastic error model of gyroscope. The gravitational vector was obtained to get horizontal attitude by compensating the outputs of accelerometers with airspeed meter, and then the attitude errors were calibrated by Kalman filter which reduced the performance requirement of gyroscope. The constraint relationship between state vector variables was regarded as pseudo-measurement equation in the measurement model, which could solve the problem of filtering divergence and singularity arisen from constraint. The adaptive adjustment of dynamic filter noise improved the accuracy of attitude estimation. The simulations and experimental results indicate that the proposed algorithm can effectively improve the system accuracy and stability.