为解决虚假目标点迹对雷达跟踪性能的影响,提出了一种基于改进K近邻(KNN)的雷达点迹真伪鉴别方法,进一步区分目标点迹和杂波点迹,滤除杂波剩余点迹,有效提高雷达处理容量和跟踪性能。该方法利用点迹形成过程中生成的特征参数,先通过核主成分分析法对特征数据降维处理,降低数据维度,提高后续算法的运行速度;再通过加权KNN算法鉴别目标点迹和杂波点迹,点迹鉴别准确率有较高提升,达到了87. 5%,算法运行速度较传统KNN算法和加权KNN算法分别提升了56%和40%。实验结果表明:该算法既有较高、较稳定的点迹鉴别准确率,又大幅度提高了算法运行速度。

In order to solve the influence of false target plot on radar tracking performance, an identification method based on improved KNN of true and false radar plot is proposed. The method further distinguishes target plot from clutter plot, filters remaining clutter plot, and effectively improves radar processing capacity and tracking performance. Using the feature parameters generated in the process of dot formation, the method first reduces the dimension of the feature data by using the core PCA to reduce the data dimension and improve the running speed of the subsequent algorithm. Then, the weighted KNN algorithm is used to identify the target plot and clutter plot, and the accuracy of plot identification is improved to 87.5%. Compared with the traditional KNN algorithm and the weighted KNN algorithm, the running speed of the algorithm has been increased by 56% and 40%, respectively. The experimental results show that the algorithm not only has a high and stable accuracy of plot identification, but also greatly improves the running speed of the algorithm.

TN957. 32