导航系统中量测滞后异步多传感器集中滤波算法

为解决异步多传感器组合存在的信息更新频率不同步,借鉴联邦滤波子滤波器的结构设计,提出在集中滤波器内根据组合时刻量测值的不同,选择相应量测矩阵进行切换的方法实现多传感器的异步组合.同时在该算法的基础上,利用非延时量测信息补偿滞后量测的方法,解决了多传感器组合量测滞后的问题.该算法在GPS/CNS/SINS组合导航实时仿真系统的应用表明,量测矩阵的切换对系统状态变量的滤波迭代没有不利影响,补偿量测滞后的方法对降低滤波误差有明显作用,滤波结果精度较高.

Concentrating filter for asynchronous multisensor with measurement delay in navigation system

The asynchronous information fusion in the multisensor system is solved by a new concentrating filter which uses the structure of the sub-federal filter as reference. In this filter the measurement matrix is a selectable unit. According to the different measurement at every integrated time, the relevant measurement matrix is selected to solve the asynchronous information fusion. In addition, the non-delayed measurement is employed to correct the delayed measurement, dealing with the problem how the delayed information is integrated. The application of this algorithm in the real-time simulated GPS/CNS/SINS integrated navigation system indicates that the matrix transition has no adverse effect on the filter iteration of the system state errors, and the method of the delayed compensation has apparent effect on the filter error reduction leading to more accurate filtering result.