基于灰关联度和距离的特征关联算法研究

A Novel Feature Association Algorithm Based on Grey Correlation Grade and Distance Measure

特征关联是无源多传感器多目标跟踪中一个关键环节。针对多传感器系统对多个辐射源进行量测过程中存在的测量误差,导致的量测值和它源于的辐射源之间出现的对应模糊问题,提出了基于灰关联度和距离度量的特征关联算法。该算法通过计算特征向量之间的灰关联度和距离,把相似度较高的特征向量划分为一类,成功地消除了对应模糊问题。模拟产生了辐射源数据库,并比较了不同相似性度量性能优劣,不同噪声环境下的关联正确率,进行了关联效果检验,仿真结果表明用灰关联度作为相似性度量关联正确率高于距离度量,但所用时间较长。

Feature association is a key link of passive multi-sensor tracking. A novel algorithm based on grey correlation grade and distance measurement is proposed to solve the problem of the corresponding fuzzy of measurement and its origination, which is caused by the measurement error produced in the process of multi-sensor system measuring the characteristic parameter of multi-emitter. The grey correlation grade and distance measure between the characteristic vectors are accounted to cluster the characteristic vectors of high similarity measurement, which eliminates the corresponding fuzzy problem successfully. Data-base of multiemitter is simulated, the performance of different similarity measurement and the correlation correct rate are discussed in different noise environments. The validity of association is proved. The simulation results show the correlation correct rate of grey correlation grade is higher than the distance measurement, but consumes longer time.