摘要
近些年来,由于水田面积迅速增加,导致三江平原井灌水稻区地下水位持续下降,吊泵、局部超采现象时有发生。为解决上述问题,以853农场为例,采用BP神经网络方法建立了853农场地下水埋深动态预测模型,对地下水埋深进行模拟和预测,精度检验结果表明,模型拟合和预测精度均较高。该模型揭示了研究区域地下水动态变化规律,为三江平原井灌区地下水资源的可持续利用提供了科学依据。
The rapid increase of paddy field acreage has led to continually descending of groundwater level in area of well irrigation paddy in Sanjiang Plain, and more and more hanging pump and partial overpumping have been occurring in area of well irrigation paddy in recent years. In order to solve above-mentioned problems, this paper took 853 Farm as example, built up the groundwater embedment depth dynamic prediction model of 853 Farm through applying the method of BP neural network. Through using this model, this paper simulated and predicted groundwater embedment depth. The results of precision test showed that the precisions of this model were high in fitting and prediction. The model built in this paper revealed the groundwater dynamic variation regularities of research area. This paper can provide scientific gist for sustainable development of groundwater resource in area of well irrigation paddy in Sanjiang Plain.
出处
《黑龙江水专学报》
2009年第1期26-30,共5页
Journal of Heilongjiang Hydraulic Engineering College
关键词
BP网络
井灌区
地下水埋深
时间序列模型
三江平原
BP network
well irrigation area
groundwater embedment depth
time-series model
Sanjiang Plain
