日光温室作物冠层温湿度时空分布及预测模型

In order to study the distribution and variation laws of temperature and humidity in the crop canopy area inside a solar greenhouse, taking the celery crop canopy in an internal heat preservation solar greenhouse in Hohhot, Inner Mongolia as the research object, the temperature and humidity at the crop canopy were tested by means of densely arranged sensors. In view of the similar situation of the variation laws of temperature and humidity at different positions of the crop canopy in the solar greenhouse, the temperature and humidity at different positions of the crop canopy were predicted through the Elman neural network. The results showed that the vertical temperature and humidity differences in the crop canopy could reach 10.24 °C and 12.97%. During the period with light (curtain opened), the differences in temperature and humidity at different positions of the crop canopy were relatively large. The temperature generally showed a distribution from high to low from top to bottom, and the humidity showed a distribution from low to high. During the period without light (curtain closed), the differences in temperature and humidity were smaller, and basically showed an opposite distribution to that during the curtain opened period. The optimized Elman neural network could predict the temperature and humidity at the crop canopy more accurately. This prediction model could predict the temperature and humidity of the crop canopy in the next week under the condition that the root mean square errors of temperature and humidity were respectively less than 0.8 and 1.5. This research has guiding significance for the monitoring and control of temperature and humidity in the crop canopy part in the solar greenhouse.

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。