基于迭代中心差分卡尔曼滤波的飞机姿态估计

在飞机姿态估计中,系统模型非线性严重、初始估计误差大和可观测性弱等固有缺陷使得对估计算法的精度要求更高。针对这一问题,给出了一种基于迭代中心差分卡尔曼滤波(ICDKF)的飞机姿态估计方法,并将其应用于由低精度高噪声传感器组成的低成本飞机姿态估计系统。首先建立基于四元数的飞机姿态数学模型,然后用ICDKF方法进行姿态估计,并通过实测数据进行验证。实验结果表明,ICDKF不仅有效地提高了飞机姿态估计的稳定性、收敛速度和滤波精度,而且无需计算雅克比矩阵,实现简单,其性能明显优于标准CDKF和EKF方法。

Iterated central difference Kalman filter based aircraft attitude estimation

In aircraft attitude estimation system, the inherent disadvantages, such as serious nonlinearity of the system mod-el, weak observability and large initial estimation error require higher estimation algorithm accuracy. Aiming at these prob-lems, a novel aircraft attitude estimation approach based on the iterated central difference Kalman filter (ICDKF) is pro-posed and applied to the low-cost aircraft attitude estimation system that contains low accuracy and high noise sensors. First, the nonlinear mathematic model of aircraft attitude based on quaternion was established. Then, the ICDKF was used for attitude estimation. The real flying data were used to verify the method. Experiment results show that ICDKF is superior to standard central difference Kalman filter (CDKF) and extended Kalman filter (EKF) methods. The algorithm not only effec-tively improves the attitude estimation precision, convergence speed and stability, but also avoids the computing burden of Jacobian matrices; and its implementation is quite simple.