一种基于模糊神经网络–模糊C均值聚类的双偏振气象雷达降水粒子分类方法

Hydrometeor Classification Method in Dual-polarization Weather Radar Based on Fuzzy Neural Network-fuzzy C-means

对于地杂波存在情况下的降水粒子分类问题,传统方法在不同的天气及环境条件下会产生较大分类误差。该文提出一种基于模糊神经网络(FNN)-模糊C均值聚类(FCM)算法的双偏振气象雷达降水粒子分类方法。该方法首先利用双偏振气象雷达在晴空模式下接收的地杂波数据训练FNN,自适应地计算地杂波各偏振参量隶属函数的参数,然后利用训练得到的地杂波隶属函数对降水模式下的地杂波进行抑制,最后采用模糊C均值聚类算法对地杂波抑制后的回波进行降水粒子分类。对实测数据的处理结果表明,该方法能够有效地抑制地杂波并获得较为精细的降水粒子分类结果。

For the problem of hydrometeor classification in the presence of ground clutter,traditional methods produce large classification errors under different weather and environmental conditions.A new method for the classification of Hydrometeor based on Fuzzy Neural Network-Fuzzy C-Means(FNN-FCM)is proposed.Firstly,the FNN is trained by the clutter data received by the Dual-polarization weather radar in the clear sky mode.The parameters of the membership function of each polarization parameter of the clutter are calculated adaptively.Then the ground clutter in the rainfall mode is suppressed by the ground clutter membership function obtained by the training.Finally,FCM clustering algorithm is used to classify the Hydrometeor after clutter suppression.The processing results of the measured data show that the proposed method can effectively suppress ground clutter and obtain finer hydrometeor classification results.