基于鲁棒自适应Kalman滤波平台标定数据处理

Processing of the Platform Calibration Data Based on Robust Adaptive Kalman Filter

惯导系统作为飞行器导航系统的重要组成部分,其精度直接影响到飞行精度。为了实现更精确的导航,通常采用射前自标定技术,从而补偿平台的误差。当飞行器竖立在发射场时,通常会受到各种干扰因素的影响,比如阵风。这些干扰会对射前标定的精度产生很大的影响。为了更好地补偿这部分干扰带来的误差,利用ARMA(2,1)建立平台漂移模型。考虑到系统噪声的方差和协方差等信息并不可知,而且所建立的模型也会有误差,选用鲁棒自适应卡尔曼滤波对数据进行处理。结果表明,经过处理后的数据方差比原数据方差小很多,可以满足标定精度要求。

Inertial navigation system as aircraft navigation system is the important component of the accuracy directly affect the accuracy of flight. In order to achieve a more accurate navigation. Calibration techniques before shoot, and compensation for the error of the platform is usually used. When the aircraft stands in launch site, it is usually affected by all kinds of interference factors, such as wind, these interference will have great affect before the precision. In order to compensation this part of the interference of the error brought better. ARMA （2, 1 ） es- tablish the platform drift model is used, considering the system of noise variance and the covariance is not known and the information such as the model also can have error. Robust adaptive Kalman filtering processing of data is choosed. The results show that after processing the data＇ s variance is lower than the original data. It can satisfy calibration precision requirements.