【作者】 胡胜； 庄旭东； 顾松山；

【Author】 Hu Sheng~1 Zhuang Xudong~2 Gu Songshan~1 ~1(Nanjing University of Information Science & Technology, Nanjing 210044, China) ~2(Guangzhou Central Meteorological Observatory, Guangzhou 510080, China)

【机构】 南京信息工程大学； 广州中心气象台；

【摘要】 设计了一种基于模糊逻辑技术的新一代雷达回波分类方法,它包含了降水回波(PDA)和奇异回波(APDA)两种探测算法。对于每种算法,在详细分析目标回波特点的基础上,提炼出一组最能描述这些回波的特征组合,如中值、平均值、标准方差、纹理结构等,并分配权重,然后利用模糊逻辑技术计算每个特征与目标回波相匹配的概率,最后对其做加权平均并选择合适阈值界定目标回波的类型。与美国WSR—88D雷达回波分类器相比,这里调整了特征量计算和联合,重新分配了权重,对PDA增加了最小反射率因子检验并加强了对反射率因子纹理结构和平均值的判断。更多还原

【Abstract】 A fuzzy logic radar echo classifying scheme for CIN-98SA is designed, and it consists of two data fusion algorithms including the AP detection algorithm(APDA) and the precipitation detection algorithm(PDA). For each algorithm, a set of features is combined to describe the characteristics of the objective echo type, and every feature is given a weight. These features include mean, median, standard deviation, texture in reflectivity, radial velocity or spectrum field. Then, the likelihood values that the features match the objective echo type are calculated by a fuzzy logic engine. By comparing the chosen threshold and the weighted average of the likelihood values, whether the observed echoes are the objective type is decided. Contrasted with the WSR-88D radar echo classifier, changes are made for the scheme including optimizing the feature combinations and adjusting the weights. Especially in the PDA, the reflectivity minimum, as a new feature is added and the weights of the reflectivity texture and mean are increased because most precipitations are convective in coastal provinces.更多还原