粒子群优化模糊聚类在信号分选中的应用

Application of PSO-FCM to Signal Sorting

针对传统的聚类算法分选正确率不高的问题,提出基于粒子群优化模糊聚类方法的分选算法,该方法利用粒子群优化算法的全局寻优能力和模糊C均值的模糊分类性质,不仅避免了梯度下降法所带来的容易陷入局部极小值的缺陷,同时也改善了不同初始聚类中心对聚类结果的影响。实验结果证明了该方法能提高雷达信号分选的正确率。

In order to improve the sorting precision of traditional clustering algorithm,a sorting al- gorithm based on particle swarm optimization fuzzy clustering method （PSO-FCM） is proposed in this paper. The method uses the seeking global excellent result performance of particle swarm opti- mization algorithm and fuzzy classification nature of fuzzy C means,which not only escapes from io- cal minimum due to gradient descent method, but also improves the influence of different initial clustering centers on clustering result. Experiment results demonstrate that the algorithm can im- prove the precision of radar signal sorting.