摘要
为了对多种关联算法进行有效集成,确保在信源系统误差估计、数据合成中使用正确关联的航迹,实现性能持续改进的信息融合系统,提出了多源航迹关联不确定度概念及其评定方法.以归一化航迹似然度为基础,综合考虑了传感器的虚情、漏情、未检测区域以及传感器参数的未确知性等因素,建立了航迹关联不确定度评定模型,并进行了仿真实验.结果表明,该模型能够正确反映目标间隔距离、传感器测量误差、航迹关联的正确性、检测概率与航迹关联不确定度之间的关系,且评定结果对传感器的系统误差、参数取值不敏感.
The concept and a method for evaluating the track correlation uncertainty were proposed to integrate several track correlation algorithms effectively, estimate the systematic errors of sensors and combine the multi-sensor data with just the right correlated tracks for constructing a self-alignment multi-sensor information fusion system. Based on the normalized track likelihood, a mathematic model for track correlation uncertainty evaluation was founded, in which the effectors, such as false positive tracks, false negative tracks, undetected area and unascertained performance of the sensors, were taken into account. Digital simulation experimental results show that the proposed method can quantitatively describe the effects of target separations, sensor measurement errors, correctness of track correlation, detection probability to the track correlation uncertainty, and evaluation is robust to the systematic errors and performance parameters of the sensors.
出处
《华中科技大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2010年第1期14-17,共4页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国防预研基金资助项目(51306030201)
关键词
数据融合
不确定度分析
评定
航迹关联
统计偏差矢量
data fusion
uncertainty analysis
evaluation
track correlation
statistical bias vector
