雹云识别的物元可拓模型及其效果检验  

Recognition Model of Hail Cloud Based on Matter Elements and Extension Sets and Its Results Verification

作　　者：李祚泳[1] 蔺雷[2] 邓新民[1] 

机构地区：[1]成都气象学院大气电子工程系,四川成都610041 [2]南京大学大气科学系,江苏南京210093

出　　处：《高原气象》2001年第2期197-201,共5页Plateau Meteorology

基　　金：广东工业大学物元分析研究所"可拓学研究项目"资助

摘　　要：雹云识别是一个多参数的模式识别问题。应用物元可拓思想将单参数雹云识别的不相容性转化为相容性问题。通过构造雹云和雷雨云两类样本的经典物元和节域物元 ,根据计算出的待识别云体与两类样本云体之间的关联度大小 ,进行雹云识别。该方法用于成都地区雹云识别结果的正确率达85Hail cloud recognition is a pattern recognition problem of multiparameter. The recognition result of hail cloud based on single parameter is always noncompatible. The basic thinking of matter clements and extension set was used for changing the noncompatible problem into compatible problem. On the basis of designing the classical matter elements matrixes and the extensional matter element matrixexes of hail clouds and thunderstorm clouds, the recognition model of hail cloud is presented according to the magnitude of correlation degrees calculated between the clouds to be recognized and the two kinds of sample clouds. The accuracy of hail cloud recognition in Chengdu by this method is over 85%.

关 键 词：雹云识别 物元 可拓集 关联函数 雷雨云 冰雹 防雹效果 

分 类 号：P426.64[天文地球—大气科学及气象学] P482