PDF全文下载：
201905017.pdf
文章编号： 16726987（2019）05011207； DOI： 10.16351/j.16726987.2019.05.017
邹丽萍1， 宫响1,2*， 庄述鹏1
(1.青岛科技大学 数理学院，山东 青岛 266061；2.中国海洋大学 环境科学与工程学院，山东 青岛 266100）
摘要： 太阳辐射的预估研究对太阳能资源的有效利用有重要意义。应用山东省福山、莒县、济南三所气象站2000—2003年的数据，建立Elman神经网络模型，对日总辐射曝辐量进行时间序列预估研究。结果表明：Elman神经网络预估效果受天气状况影响较大，晴好天气下日总辐射预估结果较精确，福山站预估与观测差值最小，范围在-2~2 MJ·m-2。城市大气污染对日曝辐量影响比较显著，模型中不考虑大气污染因素，污染较重的济南市预估效果最差，平均百分比误差变大了20%，均方根误差变大7%。Elman神经网络模型预估结果优于广义回归神经网络模型结果，3个站平均百分比误差降低5%~18%，均方根误差平均减小了0506 MJ·m-2。Elman神经网络模型适应于山东省日总辐射曝辐量的长时间预估。
关键词： Elman神经网络； 日总辐射曝辐量； 大气污染； 气溶胶光学厚度
中图分类号： TP 181文献标志码： A
引用格式： 邹丽萍， 宫响， 庄述鹏. 基于Elman神经网络的日总辐射曝辐量预估\[J\]. 青岛科技大学学报（自然科学版）， 2019， 40（5）： 112118.
ZOU Liping, GONG Xiang, ZHUANG Shupeng. Estimate of daily irradiation exposure of global radiation using Elman neural network\[J\]. Journal of Qingdao University of Science and Technology（Natural Science Edition), 2019, 40(5)： 112118.
Estimate of Daily Irradiation Exposure of Global Radiation
Using Elman Neural Network
ZOU Liping1, GONG Xiang1,2, ZHUANG Shupeng 1
(1.College of Mathematics and Physics, Qingdao University of Science and Technology, Qingdao 266061， China;
2. College of Environmental Science and Technology, Ocean University of China, Qingdao 266100， China)
Abstract: Using the data of three meteorological stations at Fushan, Juxian and Jinan, 2000—2003, the Elman neural network model was established to estimate the daily total radiation exposure. The results of Elman neural networkwas greatly affected by the weather conditions, and the estimated daily total radiation exposure under the clear weather was more accurate than that in other weathers. The difference between the estimated and observed daily total radiation exposure was smallest at Fushan stations among the three meteorological stations, ranging from -2 to 2 MJ·m-2. The urban air pollution has a significant impact on the daily total radiation exposure. If not considering the factors of air pollution in our model, both mean percentage error and root mean square error increased, especially at Jinan station (by 20%, 7%, respectively). Our results also showed that, compared with the the generalized regression neural network model, the average percentage error decreased by 5%—18%, and the root mean square error decreased by 0506 MJ·m-2 on average at the three meteorological stations. The Elman neural network model is better to estimate the daily total radiation exposure than the generalized regression neural network model. Elman neural network model is suitable for Shandong province radiation exposure to longterm projections.
Key words: Elman neural network; daily total radiation exposure; air pollution; aerosol optical depth
收稿日期： 20181009
基金项目： 国家自然科学基金项目（41406010）
作者简介： 邹丽萍（1992—），女，硕士研究生.*通信联系人.
