用于组合测姿的Kalman渐消因子自适应估计算法

提出了适用于姿态测量的Kalman滤波渐消因子自适应估计算法,滤波中采用序贯处理的方法计算出每个量测量对应的渐消因子,在位置速度组合导航系统中,只有位置、速度的误差状态是直接可观测的,用序贯滤波处理计算得到的渐消因子对协方差阵中对应于位置和速度误差状态的对角元素进行自适应控制,抑制滤波发散,提高位置、速度和姿态的测量精度。半实物仿真表明,与原来的算法相比,修改后的方法不仅能够提供高精度位置、速度信息,而且还可以提供高精度姿态信息,其中航向误差在0.08°以下,俯仰和横滚误差在0.02°左右。

Adaptive Estimation Kalman Filtering with Fading Factor for Attitude Determination in Integrated Navigation System

A method of adaptive estimation Kalman filtering with fading factor for attitude determination was proposed.Fading factors related to observations was calculated using sequential filtering method.Position’s and velocity’s error are only directly observable in integrated system by P-V integration.According to increase the weight of new observations,adaptive control using the real-time fading factors calculated by sequential filter to elements pertinent to the position’s and velocity’s error in covariance matrix’s diagonal can effectively filter the divergence and greatly improve the accuracy of integrated navigation system.Semi-physical simulation results show that the new method can accurately survey vehicle’s position,velocity and attitude.Using the new method,the error of azimuth is below,and the error of pitch and roll are about 0.02° in semi-physical simulation.