Study of ensemble forecasting based on different meteorological numerical products by using RBF artificial neural network(RBF神经网络集成多个气象数值产品的方法研究)
对MM5V3模式中16组组合区域气候模拟试验所得的结果，分区域以RBF神经网络进行集成，分析了网络结构参数的不同对集成结果的影响，最后对整个区域的温度场进行了集成.结果表明：选择合适的网络结构参数是至关重要的，神经网络的非线性集成结果均明显优于单个参数化方案和16个方案线性集成的预测结果，整个区域的温度场集成结果比MM5模拟更接近于实况场，均方根误差场也较原MM5方案明显减小.这些结果对于区域气候的集成预测以及进一步优化神经网络的性能有一定的参考意义....
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Zhejiang University Press
2006-03-01
|
| Series: | Zhejiang Daxue xuebao. Lixue ban |
| Subjects: | |
| Online Access: | https://doi.org/zjup/1008-9497.2006.33.2.223-230 |
| _version_ | 1797236314810089472 |
|---|---|
| author | GAOLing-hua(高领花) SUBing-kai(苏炳凯) TANGjian-ping(汤剑平) |
| author_facet | GAOLing-hua(高领花) SUBing-kai(苏炳凯) TANGjian-ping(汤剑平) |
| author_sort | GAOLing-hua(高领花) |
| collection | DOAJ |
| description | 对MM5V3模式中16组组合区域气候模拟试验所得的结果，分区域以RBF神经网络进行集成，分析了网络结构参数的不同对集成结果的影响，最后对整个区域的温度场进行了集成.结果表明：选择合适的网络结构参数是至关重要的，神经网络的非线性集成结果均明显优于单个参数化方案和16个方案线性集成的预测结果，整个区域的温度场集成结果比MM5模拟更接近于实况场，均方根误差场也较原MM5方案明显减小.这些结果对于区域气候的集成预测以及进一步优化神经网络的性能有一定的参考意义. |
| first_indexed | 2024-04-24T17:01:53Z |
| format | Article |
| id | doaj.art-eef831bb32ec4b4c971700f554ba9745 |
| institution | Directory Open Access Journal |
| issn | 1008-9497 |
| language | zho |
| last_indexed | 2024-04-24T17:01:53Z |
| publishDate | 2006-03-01 |
| publisher | Zhejiang University Press |
| record_format | Article |
| series | Zhejiang Daxue xuebao. Lixue ban |
| spelling | doaj.art-eef831bb32ec4b4c971700f554ba97452024-03-29T01:58:23ZzhoZhejiang University PressZhejiang Daxue xuebao. Lixue ban1008-94972006-03-01332223230zjup/1008-9497.2006.33.2.223-230Study of ensemble forecasting based on different meteorological numerical products by using RBF artificial neural network(RBF神经网络集成多个气象数值产品的方法研究)GAOLing-hua(高领花)0SUBing-kai(苏炳凯)1TANGjian-ping(汤剑平)2Department of Atmospheric Sciences, Nanjing University, Nanjing 210093, China(南京大学大气科学系，江苏 南京 210093)Department of Atmospheric Sciences, Nanjing University, Nanjing 210093, China(南京大学大气科学系，江苏 南京 210093)Department of Atmospheric Sciences, Nanjing University, Nanjing 210093, China(南京大学大气科学系，江苏 南京 210093)对MM5V3模式中16组组合区域气候模拟试验所得的结果，分区域以RBF神经网络进行集成，分析了网络结构参数的不同对集成结果的影响，最后对整个区域的温度场进行了集成.结果表明：选择合适的网络结构参数是至关重要的，神经网络的非线性集成结果均明显优于单个参数化方案和16个方案线性集成的预测结果，整个区域的温度场集成结果比MM5模拟更接近于实况场，均方根误差场也较原MM5方案明显减小.这些结果对于区域气候的集成预测以及进一步优化神经网络的性能有一定的参考意义.https://doi.org/zjup/1008-9497.2006.33.2.223-230神经网络集成预测网络结构参数区域集成 |
| spellingShingle | GAOLing-hua(高领花) SUBing-kai(苏炳凯) TANGjian-ping(汤剑平) Study of ensemble forecasting based on different meteorological numerical products by using RBF artificial neural network(RBF神经网络集成多个气象数值产品的方法研究) Zhejiang Daxue xuebao. Lixue ban 神经网络 集成预测 网络结构 参数 区域集成 |
| title | Study of ensemble forecasting based on different meteorological numerical products by using RBF artificial neural network(RBF神经网络集成多个气象数值产品的方法研究) |
| title_full | Study of ensemble forecasting based on different meteorological numerical products by using RBF artificial neural network(RBF神经网络集成多个气象数值产品的方法研究) |
| title_fullStr | Study of ensemble forecasting based on different meteorological numerical products by using RBF artificial neural network(RBF神经网络集成多个气象数值产品的方法研究) |
| title_full_unstemmed | Study of ensemble forecasting based on different meteorological numerical products by using RBF artificial neural network(RBF神经网络集成多个气象数值产品的方法研究) |
| title_short | Study of ensemble forecasting based on different meteorological numerical products by using RBF artificial neural network(RBF神经网络集成多个气象数值产品的方法研究) |
| title_sort | study of ensemble forecasting based on different meteorological numerical products by using rbf artificial neural network rbf神经网络集成多个气象数值产品的方法研究 |
| topic | 神经网络 集成预测 网络结构 参数 区域集成 |
| url | https://doi.org/zjup/1008-9497.2006.33.2.223-230 |
| work_keys_str_mv | AT gaolinghuagāolǐnghuā studyofensembleforecastingbasedondifferentmeteorologicalnumericalproductsbyusingrbfartificialneuralnetworkrbfshénjīngwǎngluòjíchéngduōgèqìxiàngshùzhíchǎnpǐndefāngfǎyánjiū AT subingkaisūbǐngkǎi studyofensembleforecastingbasedondifferentmeteorologicalnumericalproductsbyusingrbfartificialneuralnetworkrbfshénjīngwǎngluòjíchéngduōgèqìxiàngshùzhíchǎnpǐndefāngfǎyánjiū AT tangjianpingtāngjiànpíng studyofensembleforecastingbasedondifferentmeteorologicalnumericalproductsbyusingrbfartificialneuralnetworkrbfshénjīngwǎngluòjíchéngduōgèqìxiàngshùzhíchǎnpǐndefāngfǎyánjiū |