BP-ANN模型在井灌水稻区地下水埋深预测中的应用  

Based on ANN model to forecast the groundwater level on the area of well irrigation rice

作　　者：张焕昭 韩军利 宋协胜 付强[3] 

机构地区：[1]黑龙江农垦水利工程建设监理咨询有限公司,黑龙江哈尔滨150090 [2]黑龙江省大兴农场水务局,黑龙江富锦156303 [3]东北农业大学水利与建筑学院,黑龙江哈尔滨150030

出　　处：《水利科技与经济》2003年第1期21-22,67,共3页Water Conservancy Science and Technology and Economy

基　　金：中国博士后科学基金(2000)

摘　　要：　利用改进的BP算法,对三江平原创业农场井灌水稻区月平均地下水埋深进行了模拟仿真,网络拟合精度与预测精度均达到满意效果。BP-ANN模型为节约地下水开采量,恢复该地区的地下水动态平衡、制订农作物优化灌溉制度,促进农业及水资源的可持续发展提供参考作用。Through applying a kind of mended BP arithmetic which have momentum learning regular, the paper simulate the monthly groundwater depth about well irrigation rice of Chuangye Farm in Sanjiang Plain Combining the artificial neural networks with the practical problem of the area of well irrigation rice, the writer build up the ANN model Through testing and forecasting, the model precision and forecasting precision are high So, the ANN model can provide some references for many aspects, such as saving groundwater resources, resuming the balance of groundwater in this area, establishing the optimum irrigation system of well irrigation rice, developing water saving irrigation and so on It can advance the agriculture and water resource to develop continuum also

关 键 词：BP-ANN模型 井灌 水稻 地下水 人工神经网络 动态平衡 灌溉制度 

分 类 号：S511[农业科学—作物学]