基于AP密度聚类方法的雷达辐射源信号识别

Existent problems and their solutions in Miaogou Iron Ore Mine

未知雷达辐射源信号识别一直是雷达对抗情报分析中的难题。针对基于密度的聚类算法在处理不均匀样本时识别率较低的缺陷,将该算法与亲和传递（AP）聚类算法结合,提出一种基于AP密度聚类的识别方法。该方法先利用AP聚类方法对数据样本进行初步聚类,再设定相关参数,运用基于密度的带有噪声的空间聚类（DBSCAN）算法进行二次聚类。相对于原样本,初始聚类结果分布具有一定的代表性,容易找到适合DBSCAN方法的参数值。测试表明该方法具有较高的识别率。

The recent operation state and existent problems in Miaogou Iron Ore Mine are explained. The measures to solve these problems such as making appropriate mining plan, enhancing management of blasting work, optimizing crushing and dry dressing system, enhancing management of equipment use, and enhancing works of preventing geological disasters are recommended.