多传感器广义系统的加权观测融合Kalman预报器  

WEIGHTED MEASUREMENT FUSION KALMAN PREDICTOR FOR THE MULTISENSOR DESCRIPTOR SYSTEM

作　　者：陈建国[1] 冉陈键[2] 

机构地区：[1]四川绵阳职业技术学院计算机科学系,绵阳621000 [2]黑龙江大学电子工程学院,哈尔滨150080

出　　处：《系统科学与数学》2013年第12期1423-1434,共12页Journal of Systems Science and Mathematical Sciences

基　　金：国家自然科学基金(61203121);黑龙江省教育厅科学技术研究项目(12531516);黑龙江大学重点实验室研究项目;黑龙江大学青年科学基金项目(QL201112)资助课题

摘　　要：对于带相关观测噪声的多传感器线性离散随机广义系统,应用最小二乘法将各个传感器的观测信息融合为一个新的观测.再采用奇异值分解方法将原广义系统转化为两个互耦的子系统,且融合的观测也转化为其中一个状态分量的观测,从而将问题转化为单传感器常规系统的状态估计问题.对于新的降阶常规子系统,提出了全局最优的加权观测融合Kalman预报器及其预报误差方差阵.通过该方法所得广义系统的加权观测融合Kalman预报器的精度比基于状态融合的Kalman预报器的精度高,也高于局部Kalman预报器的精度,且克服了在状态融合方法中要求计算各个局部预报器的估计误差互协方差的局限性.一个多传感器广义系统的数值仿真例子说明其有效性.For the multi-sensor descriptor system with correlated measurement noises,the new fused measurement can be obtained by least square method based on all measurements.The state equation of the descriptor system can be decomposed into two normal mutual coupling sub-systems by using singular value decomposition.And the fused measurement equation is changed to the measurement of one state component.Then,the problem is the state estimation of one-sensor normal system.For the new reduced-order sub-systems,the weighted measurement fusion Kalman predictor is presented,by applying Kalman filtering.The prediction error variances of the weighted measurement fusion Kalman predictor for the descriptor system is obtained.This method avoids to compute the cross-variance of all local predictors,and the accuracy of fused predictor is higher than that of local predictors and state fusion Kalman predictor.A simulation example of 4-sensor descriptor system verifies the effectiveness of the method.

关 键 词：多传感器观测融合 广义系统 相关观测噪声 奇异值分解 Kalman预报器 

分 类 号：N941.1[自然科学总论—系统科学]