一种基于点集匹配的反舰导弹目标选择方法

摘要

In search phase of long-range anti-ship missile (ASM ) guidance, an aiming point scheme is widely used to acquire the correct target ship. However, it' s greatly influenced by the terminal dispersion of previous self control phase and the maneuvering of the target ship. In this paper, a method based on point set matching used in geometric hashing method and modified Hausdorff distance (M-HD) is presented to select a designated target from a ship formation. First, the method is used to describe point sets in geometric hashing, and the matching of two point sets under rigid transformation can be achieved by computing the distances of some pairs of fixed point sets. Second, the distance of a pair of fit point sets is measured by direct M-HD. Finally the correspondent points are acquired under bidirectional nearest neighbor rule. The method studied in this paper holds the translational invariance and rotational invariance, so its performance is not influenced by navigation error and the consistent maneuvering of the ship formation. Simulation shows the efficiency of this method and the probabilities to select the designated targets are higher than those of the traditional method using the aiming point.%远距离反舰导弹在搜索阶段选择预定目标时,目前采用的瞄准点方法受自控终点的散布误差和目标机动影响较大.为此提出用基于几何散列法和改进Hausdorff距离(M-HD)的点集匹配方法选择编队预定目标.首先利用几何散列法中描述点集的方法,将刚体变换关系下的点集匹配变成固定位置的一些点集对之间的距离计算；然后采用单向M-HD描述两个固定点集的距离；最后采用双向最近邻准则来获取对应点.该方法具有平移和旋转不变性,因此其性能与导航误差和编队的一致运动无关.仿真实验验证了该方法的有效性,且正确选择预定目标的概率高于传统的瞄准点选择方法.