相关观测噪声的加权观测融合UKF滤波器  

Weighted measurement fusion algorithm for nonlinear unscented Kalman filter

作　　者：李云[1] 郝钢[2] 邢宗新[3] 李世军 梁伯虎[1] 

机构地区：[1]哈尔滨商业大学计算机与信息学院,哈尔滨150028 [2]黑龙江大学电子工程学院,哈尔滨150001 [3]哈尔滨商业大学科技处,哈尔滨150028 [4]齐齐哈尔市邮政局计算机中心,黑龙江齐齐哈尔161000

出　　处：《哈尔滨商业大学学报（自然科学版）》2012年第3期300-303,共4页Journal of Harbin University of Commerce：Natural Sciences Edition

基　　金：黑龙江省自然科学基金(F201015)

摘　　要：对于具有相同观测方程,相关观测噪声的非系统,应用无迹卡尔曼滤波器(UKF),以及加权最小二乘(WLS)法,提出了加权观测融合UKF滤波算法.该算法具有全局最优性,且没有增加观测系统的维数,进而没有增加系统的计算负担.一个带有相关观测噪声的两传感器非线性系统的仿真例子说明了该融合算法的有效性及等价性.For nonlinear systems with the same observation equation and the same observation noise, based on the unscented Kalman filter （UKF） and weighted least squares （WLS） method, the algorithm of the weighted measurement fusion UKF was presented. The weighted measurement fusion UKF has global optimality, and it was proved that the dimension of observing system didn＇ t increase, further the algorithm, further algorithm didn＇ t increase the system＇s computational burden. A simulation example for the nonlinear systems with two sensors showed the effectiveness of the two measurement fusion UKF and verifies the com- pletely numerically identity.

关 键 词：相关观测噪声 无迹卡尔曼滤波器 加权观测融 

分 类 号：O211.64[理学—概率论与数理统计]