基于神经网络的农业灌溉量预测

以农业灌溉量为研究对象,将其看作灌区有效灌溉面积、年降雨量、粮食总产量的非线性函数,并利用神经网络和灰色预测等方法对其进行预测;在此基础上,再利用神经网络的非线性特征对灌区的灌溉供水量进行预测,并且对山东聊城彭楼灌区的数据进行了验证。结果表明:相对误差较小,预测精度达到了要求。

Forecasting Agricultural Irrigation Based on Nerve Network

The agricultural irrigation quantity to be a research object,make it as a nonlinear function of effective irrigation area,yearly rainfall and grain output,and forecast it by using nerve network and gray forecasting; On the basis of it,nonlinear function of nerve network is used for forecasting irrigation supply in irrigation area,and examine the data of Penglou irrigation area in Liaocheng of Shandong.The result is that reative error is small,forecast precision reach the demand.