摘要
根据冻融期气象资料和土壤蒸发量实测资料,利用灰色关联分析与BP神经网络相结合的方法,对冻融期大田土壤蒸发量进行了模拟预报。采用灰色关联度方法分析了影响冻融土壤蒸发的9个因子的关联度,确定了降水量、日平均气温、水面蒸发量、地表土壤温度和地表土壤含水率5个主要因子,并将其作为冻融土壤蒸发量预报模型的输入层进行模拟预测。结果表明:模型预测值与实测值的平均相对误差为9.9078%,决定系数为0.93,所建模型合理可行,可较好地用于冻融土壤蒸发预报。
Based on meteorological data and field measurement data during freeze-thaw period, a combined approach of gray correlation analysis and BP neural network model is used to simulate the freeze-thaw soil evaporation. The correlation between 9 factors which affect the evaporation of freeze-thaw soil is analyzed based on the method of gray correlation analysis. Five domain factors influencing the freeze-thaw soil evaporation, namely precipitation, average air temperature, water surface evaporation, soil surface temperature and soil surface moisture content, are determined and used as the input layer of the model to simulate and predict the evaporation of freeze-thaw soil. The results show that the mean relative error of the model between predicted and measured value is 9.9078%, the coefficient of determination is 0.93, which suggests that the Gray Correlation Analysis-BP model is reasonable and can be used for prediction of the freeze-thaw soil evaporation.
作者
解雪
陈军锋
郑秀清
薛静
高旭光
冯慧君
XIE Xue;CHEN Jun-feng;ZHENG Xiu-qing;XUE Jing;GAO Xu-guang;FENG Hui-jun(College of Water Resources Science and Engineering,Taiyuan University of Technology,Taiyuan 030024,China)
出处
《节水灌溉》
北大核心
2019年第4期22-26,共5页
Water Saving Irrigation
基金
国家自然科学基金面上项目(41572239)
国家自然科学青年基金项目(41502243)
中国博士后科学基金(2017M620098)
山西省自然科学青年基金项目(2015021169)
关键词
冻融期
土壤蒸发
灰色关联分析
BP神经网络
freeze-thaw period
soil evaporation
gray correlation analysis
BP neural network
