基于UKF的动力调谐陀螺随机漂移建模研究

随机漂移与陀螺结构及其工作环境存在非线性关系.文中对动力调谐陀螺(DTG)漂移时间序列采用AR模型进行参数粗估计,根据所建立的参数模型,采用UKF滤波方法进行滤波处理,由于UKF是一种较好的非线性滤波方法,它既消除了非线性对陀螺漂移的影响,同时克服了AR模型参数估计过程中参数不稳所带来的误差.实验结果表明,与传统的Kalman滤波方法相比,UKF对于陀螺漂移的随机项滤除效果更好.

UKF-Based Modeling Research for Dynamically Tuned Gyroscope Random Drift

The gyroscope drift is nonlinear and relative to gyroscope configuration and its operation environment. AR model is firstly applied to the DTG time series for parameter estimation. Based on the coarse parameter model, the paper proposes UKF to deal with nonlinear problem of gyroscope drift. Because of its good nonlinear filtering ability, the random drift of gyroscope drift can be greatly eliminated, and the error from parameter estimation can also be diminished. The simulation results show the filtering performance of UKF is quite satisfactory.