利用多尺度与分布模型的GPS/SINS组合导航算法

GPS/SINS Integrated Navigation System Based on Multi-scaleand Distributed Model

根据多尺度分析的思想,以离散小波变换为工具,将GPS量测信息用并行小波进行分解处理以得到不同尺度上的量测信息;然后,将不同尺度上的量测信息与SINS进行并行Kalman滤波,并对每个Kalman滤波器的输出结果进行分布式状态融合;最后,用分布式状态融合后的结果对SINS反馈校正。实际的静态实验与理论动态仿真实验结果表明,采用该方法以后,组合导航系统的各项导航参数精度明显提高。

Based on the idea of multi-scale analysis,we use the discrete wavelet transformation as a tool to decompose the GPS measurement information by the way of parallel Kalman filter,so that the measurement information in different scale can be obtained.Then,the information in different scale is integrated with SINS by the way of parallel Kalman filter,and the output results of every Kalman filter are fused by the way of distributed state estimate.At last,the results after distributed state fusion are feedback to correct SINS.The above method is validated by the practical static experiment and dynamic theoretical simulation experiment.The experimental results show that when this method is used,the precision of every navigation parameter is improved obviously.