多传感器量测噪声对航迹融合性能影响分析

Analysis of the Influence of Multisensor Measurement Noises on the Performance of Track-to-Track Fusion

针对多传感器量测噪声相关性和量测噪声方差比值对航迹融合性能有影响的问题,分别采用凸组合航迹融合、最优航迹融合和小航迹融合算法进行融合,对融合性能和单个传感器滤波性能进行分析对比。仿真实验表明:(1)当量测噪声相关系数趋向-1时,融合性能良好;当量测噪声相关系数趋向1时,融合性能退化。(2)当量测噪声方差比值大致处于[0.1,10]时,融合性能良好;当量测噪声方差比值超出此范围时,融合性能退化。

Aimed at the problem that the correlation coefficient of multisensor measurement noises and the ratio of measurement noises variance are able to influence the performance of track-to-track fusion. Convex combination fusion, optimal fusion and tracklet fusion algorithms are used to compare the state vector error between the results of fusion and single sensor filtering. Simulation results show： （1） the performance is good as the correlation coefficient tending to-1, and the performance is poor as the correlation coefficient tending to 1. （2） The performance is good as the ratio of measurement noises variance in [0.1, 10] approximately, and the performance is poor as the ratio beyond this range.