期刊文献+
共找到4篇文章
< 1 >
每页显示 20 50 100
基于LSTM-SAFCN模型的生物质锅炉NO_(x)排放浓度预测 被引量:1
1
作者 何德峰 刘明裕 +2 位作者 孙芷菲 王秀丽 李廉明 《高技术通讯》 CAS 北大核心 2024年第1期92-100,共9页
针对生物质锅炉燃烧过程的动态特性,提出一种改进的长短期记忆-自注意力机制全卷积神经网络(LSTM-SAFCN)模型用于预测NO_(x)排放浓度。首先利用完全自适应噪声集合经验模态分解法(CEEMDAN)对数据进行预处理,消除数据噪声对NO_(x)排放浓... 针对生物质锅炉燃烧过程的动态特性,提出一种改进的长短期记忆-自注意力机制全卷积神经网络(LSTM-SAFCN)模型用于预测NO_(x)排放浓度。首先利用完全自适应噪声集合经验模态分解法(CEEMDAN)对数据进行预处理,消除数据噪声对NO_(x)排放浓度预测的影响;其次融合自注意力机制与长短时记忆-全卷积神经网络(LSTM-FCN)进行特征提取与预测建模,该拓展方法能够同时兼顾时间序列数据的局部细节与长期趋势特征;最后,利用生物质热电联产系统的实际运行数据验证了所提算法的有效性。 展开更多
关键词 生物质锅炉 NO_(x)排放浓度预测 经验模态分解 长短时记忆-全卷积神经网络(LSTM-FCN) 自注意力机制
下载PDF
火电厂大气污染物排放预测模型 被引量:1
2
作者 张冉 张山山 +1 位作者 史一涛 王诗琴 《环境工程学报》 CAS CSCD 北大核心 2016年第5期2547-2550,共4页
根据大气污染物排放浓度变化特点,将无偏GM(1,1)模型与神经网络模型组合,并以矩阵型输入方式替代传统的数列型数据输入方式,得到改进型灰色神经网络模型,称为UGMN模型。接着,采用烟囱入口烟气自动监控系统(CEMS)数据,将模型运用于贵州... 根据大气污染物排放浓度变化特点,将无偏GM(1,1)模型与神经网络模型组合,并以矩阵型输入方式替代传统的数列型数据输入方式,得到改进型灰色神经网络模型,称为UGMN模型。接着,采用烟囱入口烟气自动监控系统(CEMS)数据,将模型运用于贵州省某电厂白天及夜间两段时间段内大气污染物排放浓度的模拟与预测。研究结果表明UGMN模型预测精度较好,可以应用于火电厂大气污染物排放浓度预测。 展开更多
关键词 灰色神经网络 无偏GM(1 1) 火电厂大气污染 排放浓度预测
原文传递
Impacts of Secondary Aerosols on a Persistent Fog Event in Northern China 被引量:6
3
作者 JIA Xing-Can GUO Xue-Liang 《Atmospheric and Oceanic Science Letters》 2012年第5期401-407,共7页
The chemistry version of the Weather Re- search and Forecasting model (WRF/Chem) was coupled with the anthropogenic emission inventory of David Streets to investigate the impacts of secondary aerosols on a persisten... The chemistry version of the Weather Re- search and Forecasting model (WRF/Chem) was coupled with the anthropogenic emission inventory of David Streets to investigate the impacts of secondary aerosols on a persistent fog event from 25 to 26 October 2007, in Northem China. The spatial distribution of the simulated fog is consistent with satellite observations, and the time-height distributions of the simulated boundary layer where the fog formed are also in good agreement with these observations. The sensitivity studies show that the secondary aerosols of SO4, NO3, and NH4 formed from gaseous precursors of SO2, NOx, and NH3 had substantial impacts on the formation processes and microphysical structure of the fog event. The decrease of the secondary aerosols obviously reduced the liquid water path and column droplet number concentration of the fog below the 1-km layer, and the corresponding area-averaged liquid water path and droplet number concentration of the fog decreased by 43% and 79%, respectively. The concentra- tions of NOx and NO3 were found to be extremely high in this case. The concentration of interstitial aerosol NO3 was much higher than the SO4 and NH4, but the concentration of SO4 was highest in the cloud-borne aerosols. The average activation ratios for SO4, NO3, and NH4 were 34%, 31%, and 30%, respectively, and the maximum ra- tios reached 62%, 86%, and 55% during the fog episode. 展开更多
关键词 secondary aerosol FOG WRF/Chem simulation
下载PDF
Projections of Global Mean Surface Temperature Under Future Emissions Scenarios Using a New Predictive Technique 被引量:1
4
作者 WANG Ge-Li YANG Pei-Cai 《Atmospheric and Oceanic Science Letters》 CSCD 2014年第3期186-189,共4页
Using numerical model simulations, global surface temperature is projected to increase by l^C to 4~C during the 21 st century, primarily as a result of increasing concentrations of greenhouse gases. In the present stu... Using numerical model simulations, global surface temperature is projected to increase by l^C to 4~C during the 21 st century, primarily as a result of increasing concentrations of greenhouse gases. In the present study, a predictive technique incorporating driving forces into an observation time series was used to project the global mean surface temperature under four representative sce- narios of future emissions over the 21st century. 展开更多
关键词 CLIMATE predictiondriving forcesprojection
下载PDF
上一页 1 下一页 到第
使用帮助 返回顶部