BP神经网络的春小麦作物水模型的初步研究

Preliminary Study on Spring Wheat Response to Water with BP Neural Network Method

用内蒙古河套灌区春小麦非充分灌溉试验资料 ,建立了 BP神经网络的春小麦作物水模型。经模拟分析认为该模型能正确表达春小麦的产量与水分的关系并有某些独特的优点。与当地拟合效果较好的Minhas模型的比较表明 ,二个模型所表达的春小麦各生育阶段对水分的敏感性一致 ,产量预测结果接近。初步说明

Based on irrigation experiment data in spring wheat in Hetao Irrigation District, a model of simulating spring wheat response to water with BP neural network method is developed. Simulating results showed that the model is reliable to describe the relationship between yield and water use in spring wheat. Compared with Minhas model, which was proved as a good model to local situation, the new model may provide nearly the same results for analyzing the yield sensibility to water application. Research results indicate that the BP neural network is a suitable method to simulate crop response to water.