欢迎访问《兵工学报》官方网站,今天是 分享到:

兵工学报 ›› 2023, Vol. 44 ›› Issue (2): 472-483.doi: 10.12382/bgxb.2021.0660

• • 上一篇    下一篇

基于Hough变换和优化K均值聚类的反舰导弹编队识别目标选择方法

黄隽*(), 吴鹏飞, 李晓宝, 刘玥   

  1. 海军工程大学 兵器工程学院, 湖北 武汉 430033
  • 收稿日期:2021-10-06 上线日期:2022-06-09
  • 通讯作者:
  • 基金资助:
    国家自然科学基金项目(62073334)

Formation Recognition and Target Selection of Anti-ship Missile Based on Hough Transform and Optimized K-means Clustering

HUANG Jun*(), WU Pengfei, LI Xiaobao, LIU Yue   

  1. College of Weaponry Engineering, Naval University of Engineering, Wuhan 430033, Hubei, China
  • Received:2021-10-06 Online:2022-06-09

摘要:

现有反舰导弹编队识别技术存在侧重于队形识别而非目标选择、对编队末端态势变化考虑不足、实时性偏弱、未考虑队列线检测区间对聚类效果的影响和聚类数优化与聚类迭代过程相互独立等问题。基于Hough变换和优化K均值聚类算法,提出反舰导弹编队识别目标选择流程,构建V形、平行和环形编队目标生成与目标选择模型,旋转、缩放、冲淡式干扰和队型变化等编队目标变化模型。仿真结果表明,聚类数优化代价函数在关键聚类数段区分度明显、聚类数优化准确,移动检测区间检测解决正常检测区间两侧边缘样本点对应机制缺失问题,采用多样本更新聚类与单样本更新聚类结合的聚类迭代、聚类数优化迭代与聚类迭代融合迭代效率高,工程适用性强,对于反舰作战模拟具有重要意义。

关键词: 反舰导弹, Hough变换, 优化K均值聚类, 编队识别, 目标选择

Abstract:

For the formation recognition technology of anti-ship missiles, there are problems of focusing on formation recognition rather than target selection, insufficient consideration of the changes from terminal guidance stage to target indication stage, and relatively weak real-time performance and so on. The formation recognition target selection technology has replaced the single feature recognition target selection technology and become the mainstream of target selection for long-range anti-ship missile weapon systems based on Hough transform and optimized K-means clustering algorithm. A process of formation recognition and target selection for anti-ship missiles is proposed, and models on targets' generation, selection and changes, such as formation rotating, scaling, transforming and dilution jamming, of V-shaped, parallel and ring formations are built up. The simulation results show that: the cost function for optimal clustering number is more effective than the existing methods; the problem for the lack of transforming mechanism for the sample points on both sides of the normal detection interval is solved using the formation line detection on the moving detection interval; the clustering iteration, optimal iteration of clustering number, and fused cluster iteration by combining multi-sample update clustering with single-sample update clustering have higher efficiency and stronger engineering applicability, which is of great significance for anti-ship combat simulation.

Key words: anti-ship missile, Hough transform, optimized K-means clustering, formation recognition, target selection

中图分类号: