作物需水量的BP神经网络预测模型研究

In this paper, a network topological structure was determined and an artificial neural network model for predicting crop water requirements was established by using BP neural network theory and correlation analysis between many factors and crop water requirements. The writer took winter wheat water requirements as export and the fallow six data as input: the average daily air temperature, relative humidity, wind speed at 2-m height, the bright sunshine hours per day, average soil water content and the winter wheat growing days/ growing degree-days/ growing factor, which were chose from the meteorological data, winter wheat data and soil water content data of Yong le dian, from the year 1998 to 2001. The BP neural network was trained by above data and tested as well by a group of non-training sample. The results show that: (1) During crop growing days, the effect of air temperature and soil humidity can not be ignored, and growing factor had residual effect or cumulative response to crop evapotranspiration. (2) The BP neural network can solve the uncertainty between crop water requirements and many factors, research results preliminary indicate that the BP neural network is a suitable method to predict crop water requirements.