摘要
为了研究日光温室内部作物冠层区域温湿度分布及变化规律,以内蒙古呼和浩特市内保温型日光温室西芹作物冠层为研究对象,采用传感器密集布点的方式测试作物冠层处温湿度,针对日光温室作物冠层不同位置温湿度变化规律相似的情况,通过Elman神经网络预测作物冠层不同位置的温湿度情况。结果表明:作物冠层垂直温湿度差可达10.24℃,12.97%。在有光照(起帘)时期,作物冠层不同位置温湿度差异相对较大,温度由上到下总体呈现从高到低、湿度由低到高的分布,在无光照(闭帘)时期则温湿度差异较小,基本与启帘时期呈现相反分布。优化后的Elman神经网络能够较准确预测作物冠层处温湿度。该预测模型可在保证温度、湿度均方根误差分别小于0.8、1.5的情况下预测未来一周的作物冠层温湿度,该研究对日光温室内作物冠层部分温湿度监测与控制具有指导意义。
In order to study the distribution and change situation of temperature and humidity inside the canopy area of solar greenhouse.Taking celery canopy of heat preservation solar greenhouse in Hohhot,Inner Mongolia as the research object,temperature and humidity distribution of celery canopy was measured and observed by means of sensor dense distribution.According to the similar rule of temperature and humidity in different positions of greenhouse canopy in solar greenhouse,prediction of temperature and humidity in different positions of crop canopy using Elman neural network.The results showed that the vertical temperature and humidity difference of crop canopy can reach 10.24 ℃,12.97%.During the period of illumination(curtain opening),the distribution of temperature from top to bottom,humidity from low to high,and in the period of no illumination(curtain closing) was basically opposite.The optimized Elman neural network could predict the temperature and humidity of crop canopy accurately.The model could predict the temperature and humidity of crop canopy in the coming week under the condition that the root mean square error of temperature and humidity were less than 0.8 and 1.5 respectively.This study has guiding significance for monitoring and controlling the temperature and humidity of crops canopy in solar greenhouse.
作者
刘琦
塔娜
焦巍
康宏源
赵志勇
LIU Qi;TANA;JIAO Wei;KANG Hongyuan;ZHAO Zhiyong(College of Mechanical and Electrical Engineering,Inner Mongolia Agricultural University,Hohhot,Inner Mongolia 010011)
出处
《北方园艺》
CAS
北大核心
2019年第17期56-65,共10页
Northern Horticulture
基金
国家自然科学基金资助项目(61663038)
关键词
作物冠层
温湿度
分布
预测模型
crop canopy
temperature and humidity
distribution
prediction model
