摘要
为得出浙江省干旱趋势和简化估算模型,以全区域9个站点为研究对象,计算不同站点的标准化降雨蒸散指数(SPEI)。同时以卷积双向长短期记忆神经网络模型(CNN-BiLSTM)为基础,采用小波包变换(WPT)优化的蜣螂算法(DBO)和珍鲹算法(GTO),构建2种优化组合模型,并比较不同模型精度,结果表明:全年春旱呈现逐渐加剧趋势,WPT-DBO-CNN-BiLSTM模型在所有模型中精度最高,可推荐用于预测全区不同尺度的SPEI。
In order to obtain the drought trend and simplified estimation model of Zhejiang Province,the standardized precipitation evapotranspiration index(SPEI)of 9 stations in Zhejiang Province was calculated,and the convolutional bidirectional long short-term memory neural network model(CNN-BiLSTM)was used as the basis.Dung Beetle Algorithm(DBO)and Jean Caranx algorithm(GTO)optimized by wavelet transform(WPT)were used to construct two optimal combination models,and the accuracy of different models was compared.The results showed that the spring drought gradually intensified throughout the year,and the WPT-DPO-CNN-BiLSTM model had the highest accuracy among all models,and could be recommended for predicting the SPEI index of different scales in the whole region.
作者
陈剑峰
周培华
CHEN Jianfeng;ZHOU Peihua(Hangzhou Huyuan Township People's Government,Hangzhou 311400,Zhejiang,China;Puyang River Basin Management Center Hangzhou Xiaoshan District,Hangzhou 311200,Zhejiang,China)
出处
《浙江水利科技》
2024年第5期54-61,共8页
Zhejiang Hydrotechnics
关键词
浙江省
标准化降雨蒸散指数
卷积双向长短期记忆神经网络
小波包变换
蜣螂算法
Zhejiang Province
the standardized precipitation l evapotranspiration index
convolutional bidirectional long short-term memory neural network
wavelet transform
Dung Beetle Algorithm