论文部分内容阅读
为了克服联邦滤波器理论中对滤波器维数的限制,基于信息滤波算法和信息分配原理,发展了广义联邦滤波器设计理论。提出了当主滤波器维数和局部滤波器维数不相同时,广义联邦滤波器达到全局最优的解析补偿方法。并将上述设计方法应用到双MIMS/GPS组合导航系统实验中。实验结果表明:经解析补偿的广义联邦滤波器的定位精度同集中Kalman滤波器的定位精度十分接近,具有全局滤波最优性。广义联邦滤波器的设计使联邦滤波理论更加完善,增强了其实用性。
In order to overcome the limitation of the filter dimension in the federal filter theory, a generalized federal filter design theory is developed based on the information filtering algorithm and information distribution principle. Proposed a generalized federal filter to achieve global optimal parsing and compensating method when the main filter dimension and the local filter dimension are not the same. The above design method is applied to the experiment of double MIMS / GPS integrated navigation system. The experimental results show that the localization accuracy of the generalized federated filter with analytical compensation is very close to the localization accuracy of the centralized Kalman filter, and it has the global filtering optimality. The design of the generalized federal filter makes the federal filter theory more perfect and enhances its practicality.