测量噪声相关情况下改进的多传感器融合算法  

An improved algorithm of multisensor fusion with correlated measurement noise

作　　者：巴宏欣[1] 奚和平[1] 杨飞[1] 赵宗贵[2] 

机构地区：[1]解放军理工大学指挥自动化学院,江苏南京210007 [2]中国电子科技集团公司第28研究所,江苏南京210014

出　　处：《解放军理工大学学报（自然科学版）》2005年第3期205-208,共4页Journal of PLA University of Science and Technology(Natural Science Edition)

基　　金："十五"国防科技预研基金资助项目(413060301);国防基金资助项目(J23-1.5).

摘　　要：为了研究测量噪声相关情况下,同步多传感器跟踪系统的测量融合技术,根据线性无偏最小方差估计理论,给出了测量噪声相关情况下,同步多传感器系统测量融合的伪序贯和数据压缩实现算法,它们对测量噪声相关与不相关情况都适用。仿真表明,在测量噪声相关时,尤其是相关性较强时,本算法的融合精度比传统算法有明显提高;而测量噪声不相关时,性能与传统的融合算法相同。仿真结果表明了算法的有效性。The measurement fusion techniques of synchronized multi-sensor track i ng system in case of correlated measurement noises were studied. Under the condi t ion of correlated measurement noises, the algorithms of pseudo-sequential filte r ing and compressing implement of measurement fusion in synchronized multi-senso r sy stem were presented based on linear unbiased minimum variance estimation theo ry. Compared with the classical multi-sensor measurement fusion algorithms, the theory analysis and simulation show that the algorithms presented in this pape r improve the estimation accuracy obviously.

关 键 词：融合算法 声相关 多传感器系统 融合技术 跟踪系统 估计理论 最小方差 实现算法 数据压缩 测量融合 传统算法 测量噪声 仿真结果 相关性 同步 

分 类 号：TP274.2[自动化与计算机技术—检测技术与自动化装置] U674.703.5[自动化与计算机技术—控制科学与工程]