本論文研究以微機電飛行資訊量測系統為基礎之飛行姿態角之計算，提出以四元素法為核心之擴展式卡曼濾波器的設計，探討整合傳統四元素法及重力場計算飛行姿態角之運算法則，並以實際飛行數據執行實務驗證。 此擴展式卡曼濾波器以傳統計算飛行姿態角之四元素法為動態模式，四個元素即為濾波器之狀態，重力場計算所得之飛行姿態角則視為系統之實際量測值。本研究首先利用四軸運動平台設計一個固定滾轉角飛行的實驗，由飛行資訊量測系統量測運動過程的三軸加速度及角速度，並分別由四元素法及重力場計算飛行姿態角，同時探討實驗誤差及量測雜訊之特性，之後再利用擴展式卡曼濾波器，做互補濾波，計算姿態角，驗證此卡曼濾波器設計的實用性。最後再以實際飛行數據來驗證本論文所提出的飛行姿態角之計算法則。 This thesis studies the attitude determination using a MEMs flight information measurement unit. A quaternion-base extended Kalman filter to integrate the traditional quaternion and gravitational methods for attitude determination algorithm is proposed in the study. The proposed extended Kalman filter utilizes the evolution of the four elements in the quaternion method for attitude determination as the dynamic model with the four elements as the states of the filter. The attitude obtained from the gravity computation is treated as the output of the filter. An experiment using an in-house designed motion platform is designed to evaluate the proposed algorithm. From the experiment, the noise characteristics of the sensor signals are examined. The proposed filter design is then validated using the laboratory data. With the success in the laboratory test, a set of flight test data is utilized to show the practicality of the filter design
