加氧灌溉温室甜瓜植株蒸腾人工神经网络分析  

Research on BP artificial neural network model of greenhouse muskmelon plant transpiration under aeration drip irrigation

作　　者：谢恒星[1,2] 蔡焕杰[2] 

机构地区：[1]渭南师范学院化学与生命科学学院,陕西渭南714000 [2]西北农林科技大学旱区农业水土工程教育部重点实验室,陕西杨凌712100

出　　处：《广东农业科学》2013年第14期44-47,共4页Guangdong Agricultural Sciences

基　　金：国家自然科学基金(50779059);陕西省教育厅科研计划项目(12JK0814);西北农林科技大学旱区农业水土工程教育部重点实验室访问学者项目与渭南师范学院研究生专项科研项目(12YKZ056)

摘　　要：应用热技术方法研究了不同天气条件下温室甜瓜植株蒸腾速率的变化特征,利用BP人工神经网络模型对植株蒸腾与环境因子的定量关系进行了模拟。结果表明,晴天条件下温室甜瓜植株蒸腾日变化基本呈对称单峰曲线,阴天和雨天条件下植株蒸腾启动时间较晚且变化和缓;不同天气条件下的BP人工神经网络模型均具有较高的拟合精度,但雨天条件下当植株蒸腾速率较高时模型的偏差较大。This paper analysed plant transpiration ratio diurnal variation of greenhouse muskmelon under different weather conditions using heat technology, discussed the relationship between plant transpiration and environmental factors using BP ＇artificial neural network model. Results showed that the diurnal variation of plant transpiration of greenhouse muskmelon presented single peak curve, the plant transpirations start point delayed and had small change in cloudy and rainy day. The BP artifieial neural network model had high precision to simulate palant transpiration on different weather conditions but the model had relatively high error in rainy day when the plant transpiration increased.

关 键 词：植株蒸腾 BP人工神经网络 蒸腾模拟 

分 类 号：S275.6[农业科学—农业水土工程]