利用模型预测地下水埋深的动态

Dynamic prediction of groundwater buried depth by using model

文中将资料记载的地下水埋深时间序列,利用双向差分和自忆性原理,建立DAMSM模型,求得预测方程,其预测结果最大相对误差仅为8.07%,平均相对误差为4.33%,拟合性优良,可用于后续的地下水埋深动态监测控制管理。该模型将数据的历史信息纳入模型计算,填补了用单因素预测的缺陷。最后,根据首山漏斗区的地质和地下水特性,提出地下水生态环境缓解和修复方案建议。

According to the groundwater buried depth time series coming from the information recordation,the DAMSM model is established and the prediction equation is obtained by using the bi-directional difference and self-memorization principle.For the prediction results,the maximum relative error is only 8.07%,the average relative error is about 4.33%,the fitting is good,which could be used to control and manage the groundwater depth dynamic monitoring.Through the history data is used in the model calculation,the model could fill the defects causing by the single prediction.At last,the alleviation and restoration scheme of groundwater ecological environment are put forward based on the geology and groundwater characteristics of Shoushan funnel area.