摘要
为找出辽宁省干旱的基本规律,并找出适用于辽宁省干旱预测的标准模型,基于该区域沈阳、锦州、开原、叶柏寿、瓦房店、宽甸、岫岩7个气象站点的气象数据,计算不同站点的相对湿润度指数,并基于极限学习机模型、M5树模型、随机森林模型共3种机器学习模型,建立辽宁省干旱预测模型,结果表明:辽宁省多年干旱程度逐年降低,且当地在春季和冬季的干旱严重,存在明显的季节性干旱现象,极限学习机模型在模拟预测干旱中表现的精度较高。
In order to find out the basic law of drought and to find a standard model suitable for drought prediction in Liaoning Province, the article is based on the Data, calculate the relative humidity index of different sites, and build a drought prediction model in Liaoning based on three machine learning models: extreme learning machine model, M5 tree model, and random forest model. And the local drought in spring and winter is severe, and there is obvious seasonal drought phenomenon. The extreme learning machine model has higher accuracy in simulating and predicting drought.
作者
林佳楠
Lin Jianan(Fuxin Water Conservancy Affairs Service Center,Fuxin 123000,Liaoning)
出处
《陕西水利》
2020年第3期54-56,共3页
Shaanxi Water Resources
关键词
辽宁省
干旱预测
极限学习机
相对湿润度指数
季节性干旱
Liaoning Province
the drought prediction
extreme learning machine
the relative humidity index and seasonal drought
