传感器位置条件的自适应卡尔曼滤波及其应用

Adaptively Kalman filtering based on the constrained conditions of position and application in mobile mapping system

当多个传感器安装于同一载体之上,构成组合导航系统时,必然存在位置约束关系,当定位传感器等于或多于三个时,会产生多个这样的关系。利用这些条件,采用传统的约束滤波算法可以提高组合定位系统的整体精度,但有时也会造成高精度传感器的精度损失。本文中将这些位置约束条件看作观测量,通过设计适当的权矩阵,并结合自适应卡尔曼滤波算法,提出了一种用于提高传感器滤波精度的方法,并和传统的约束滤波算法进行了比较。仿真计算表明:在各传感器精度相当时,该算法可以提高各个传感器的精度,并和传统的约束滤波算法等效;当各传感器精度不同时,该算法仍然可以提高高精度传感器的滤波精度或保证高精度传感器的滤波精度不受损失。

Traditional constrained filter can improve the precision of the integrated position system,but it will possibly lose the accuracy of high sensors.Regarding constrained conditions as observations and designing a right matrix,this paper presented an algorithm combining with the adaptively Kalman filter to make a comparison with the traditional adaptively constrained filter.The simulated example showed：when the precisions of the sensors are similar,the algorithm could improve the precision for each sensor,and it is equivalent with traditional adaptively constrained filter;while the precisions are different,the algorithm could still improve or at least keep the filter precisions of high sensors.