基于粗糙集方法的暴雨灾害分析与预报  

Analysis and Forecast of Rainstorm Disaster Based on Rough Set Method

作　　者：单浠 朱天心 叶佳欣 王燚潇 孙佩茹 陈雪东[1] SHAN Xi;ZHU Tianxin;YE Jiaxin;WANG Yixiao;SUN Peiru;CHEN Xuedong(School of Science,Huzhou University,Huzhou 313000,China)

机构地区：[1]湖州师范学院理学院,浙江湖州313000

出　　处：《湖州师范学院学报》2021年第2期21-27,共7页Journal of Huzhou University

基　　金：国家自然科学基金项目(11171105)。

摘　　要：以上海奉贤地区1959—2019年暴雨相关数据为例,首先填补缺失数据,根据粗糙集理论和方法对气压、气温、湿度、风速、降水5个属性采用3种不同的方法离散化处理,择优选取最佳离散方法,再加以属性约简后建立数据库,并用神经网络计算发生暴雨的概率,为国家提前预测暴雨气候提供有益参考.Taking the rainstorm data from 1959 to 2019 in Fengxian area of Shanghai as an example,this paper research on historical data,establish a rainstorm knowledge base,and provide data support for the prediction of rainstorm disasters.We fill the missing data at first.According to the rough set theory and method,the five attributes of air pressure,temperature,humidity,wind speed and precipitation are discretized by three different methods.Choose the best dispersion method,and then perform attribute reduction.We establish the database,and calculate the probability of the rainstorm by BP neural network.This research provides a helpful reference for the prediction of the country’s torrential rain climate.

关 键 词：粗糙集 离散分类 属性约简 知识库 天气预报 

分 类 号：F832[经济管理—金融学]