摘要
根据山东簸箕李灌区冬小麦灌溉实测资料,初步建立了基于BP网络的畦灌性能模拟模型,并以地面灌溉数值模型SRFR模拟结果作为目标值,对开发的BP模型进行了训练,确定了BP模型运行参数。应用结果表明,SRFR模型和BP模型得到的灌水效率和灌水均匀度变化规律一致,平均相对误差均为2.5%左右,基于BP网络的畦灌性能模拟模型可用于类似条件下畦灌性能指标的预测和评价。
Based on field experiment data in Bojili Irrigation District, Shandong province, a border irrigation performance simulation model with application of the BP neural network is developed. Based on the results from the surface irrigation numerical simulation model SRFR, the BP model is trained and the parameters needed to run the BP model are determined. The application of the BP model shows that the simulated efficiency of the application Eo and the distribution uniformity Du from two models have similar variation tendency. The average relative errors between the results from two models are about 2.5%. The developing border performance simulation model with application of the BP neural network could be used for foreseeing and evaluations on performance of the border irrigation system under similar conditions.
出处
《灌溉排水学报》
CSCD
北大核心
2007年第6期47-50,共4页
Journal of Irrigation and Drainage
基金
"十五"国家重大科技专项(863计划)课题(2002AA2Z4041)
关键词
BP神经网络
畦灌
灌水效率
灌水均匀度
BP neural network
border irrigation
irrigation efficiency
distribution uniformity
