基于气象预报的参考作物蒸发蒸腾量的神经网络预测模型

Neural network model for reference crop evapotranspiration prediction based on weather forecast

参考作物蒸发蒸腾量(ET0)是进行实时灌溉预报和农田水分管理的主要参数,BP神经网络能够较好地反映ET0与诸影响因素间复杂的非线性关系。本文将ET0看作时间序列,选取前3日ET0作为影响因子,以天气预报可测因子包括最高、最低和日平均温度、反映天气类型的阴晴指数、日序数和风力等级进行修正,建立了三层BP神经网络模型。选取江苏射阳站2003与2004年气象资料,应用Matlab神经网络工具箱,采用trainer算法进行模型训练与预测。结果证明,所建模型能够很好地反映诸多影响因子与ET0之间的关系,具有较高的模拟精度和较好的泛化能力。

The variation process of reference crop evapotranspiration（ ETo ）was regarded as time series and the ETo data in passed three days were selected as the influencing factors to establish a BP neural network model with three layers for prediction. The meteorological data obtained from weather forecast, such as maximum, minimum and average daily temperature, weather index, number of days in the year and windforce, were used to improve the performance of the model. The observation data from Sheyang County, Jiangsu Province, were used to train and test the model. The result shows that the proposed model can well reflect the relationships between ETo and relevant factors.