基于目的地预测的多相似加权目标编群方法

针对目标编群中单一算法存在的适用范围小、误分率高的问题,提出一种新的态势估计中目标编群的处理方法。首先应用Hop fie ld神经网络对态势中目标的目的地做出判断,然后采用多相似性加权策略计算出目标间的相关系数,再根据最大相关系数层次聚类算法实现编群。仿真结果表明方法能在一定程度上减小错误编群的概率,同时适用范围也得到了扩展。

A Target Grouping Method with Weighted Similarity Measure Based on Pre-destination

Towards the limited range and high error rate of the traditional target grouping method,we present a novel model to deal with the target grouping problem.Firstly,A Hopfield neural network is selected to resolve the pre-destination of the targets;then,the weighted similarity measure is used to calculate the relative value;thirdly,the target clustering is implemented by the hierarchical clustering algorithm.The result indicates that the method can reduce the error rate of clustering and enlarge the available ...