多传感器ARMA信号加权融合Kalman滤波器  

Weighted fusion Kalman filter for multi-sensor ARMA signal with measurement delays

作　　者：封文清[1] 杜鑫龙[1] 孙赫[1] 邓自立[1] 

机构地区：[1]黑龙江大学电子工程学院,哈尔滨150080

出　　处：《黑龙江大学自然科学学报》2013年第4期458-461,共4页Journal of Natural Science of Heilongjiang University

基　　金：国家自然科学基金资助项目(60874063)

摘　　要：对带观测滞后和有色观测噪声的多传感器自回归滑动平均(ARMA)信号,提出一种反卷积加权融合Kalman滤波方法,其特点是将信号滤波问题转换为反卷积估计问题,并将观测滞后嵌入到反卷积模型中。仿真例子说明,按标量加权融合算法的精度高于每个局部滤波器的精度,说明了算法的有效性。For multi-sensor ARMA signal with the measurement delays and colored noises, a deconvolution weighted fusion Kalman filtering method is presented where the signal filtering problem is converted into a deconvolution esti- mation problem and the measurement delays are embedded into the deconvolution model. Simulation example illus- trates that the accuracy of the fused algorithm weighted sensor by scalars is higher than that of each local filter. A simple example shows its effectiveness.

关 键 词：多传感器信息融合 时滞 有色噪声 反卷积Kalman滤波方法 

分 类 号：O241.64[理学—计算数学]