CAJ | 학술논문

针对扩展卡尔曼滤波方法在飞机姿态滤波中存在的线性化误差大、需要解繁琐的Jacobian矩阵等问题，将一种新型卡尔曼滤波方法--单向无迹卡尔曼滤波应用于载有低精度、高噪声传感器的低成本飞机姿态滤波系统，并在精度及计算量上与EKF和容积卡尔曼滤波进行了比较。利用实测飞行数据进行实验，结果表明：相对于EKF，SUKF实现容易，且使姿态滤波精度提高到二阶；相对于CKF， SUKF计算简单，比CKF减少约33%的计算量。
침대확전잡이만려파방법재비궤자태려파중존재적선성화오차대、수요해번쇄적Jacobian구진등문제，장일충신형잡이만려파방법--단향무적잡이만려파응용우재유저정도、고조성전감기적저성본비궤자태려파계통，병재정도급계산량상여EKF화용적잡이만려파진행료비교。이용실측비행수거진행실험，결과표명：상대우EKF，SUKF실현용역，차사자태려파정도제고도이계；상대우CKF， SUKF계산간단，비CKF감소약33%적계산량。
In view that the Extended Kalman Filter(EKF) in aircraft’s attitude filtering has such problems as large linearization error and having to solve complex Jacobian matrix, etc., a new Kalman filter algorithm named Simplex Unscented Kalman Filter(SUKF) is applied into a low-cost aircraft attitude filtering system equipped with less accurate and high noisy sensors. Meanwhile, a comparison on accuracy and calculation cost among EKF, Cubature Kalman filter(CKF) and SUKF is conducted. Experimental results with real flying data show that the SUKF is easy to implement and increases the attitude filtering accuracy to the second order with respect to EKF. Moreover, the SUKF is simple in computation, and the computational cost decreases about 33%compared with that of CKF.