基于观测值聚类的多雷达数据融合

Radar data fusion based on clustering measurements

根据多部雷达同一时刻对同一目标的观测值在空间呈团状的特征,运用模式识别理论中聚类的方法解决数据融合问题。采用一种改进的KNN算法对多雷达观测数据进行聚类,结合聚类中心和目标预测值,应用卡尔曼滤波器估计目标状态,从而实现多雷达数据融合。实验结果表明,这种方法是有效的。

The feature of the measurements of the same target at the same time makes it possible to use clustering technique to solve data fusion problem. This paper uses an improved K-nearest neighbor algorithm to cluster the data from radars and intergrates the centers of the clusters with premeasurements, then estimates the state of targets with Kalman filter, thus realize the data fusion. The result of the experiment shows that the method is effective.