基于多尺度预处理的GPS/SINS组合导航系统

GPS/SINS Integrated Navigation System Based on Multi-scale Preprocessing

基于多尺度分析的思想,以离散小波变换为工具,利用小波对惯性元件输出的信息进行并行阈值消噪以削弱惯性元件误差对SINS及组合系统性能的影响;然后,对GPS输出的信息进行并行多尺度预处理,并结合传统的Kalman滤波方法,对系统进行综合滤波;将上述方法引入到GPS/SINS组合导航系统中,利用实测数据进行验证,并给出了基于不同方法的大量实验曲线。实验结果表明,该方法可以有效削弱惯性元件以及GPS误差对系统的影响,提高了GPS/SINS组合导航系的精度和可靠性。

In the GPS/SINS integrated navigation system,the precision of inertial elements and GPS output is the important factor which affects the precision of integrated navigation system.Thus,based on the idea of multi-scale analysis,this paper uses wavelet to de-noise in parallel the inertial elements in order to abate the effect of inertial element on SINS and integrated navigation system;then the parallel multi-scale process is introduced to GPS output,and the Kalman filtering algorithm based on measurement multi-scale preprocessing is given combining the traditional Kalman filter;at last,the above method is introduced to the GPS/SINS integrated navigation system,the measured data is used to verify the above method,and many experimental curves are given based on different methods.It is shown that this method can effectively decrease the effect of error of inertial elements and GPS output on the overall system and improve the precision and reliability of navigation.