摘要
根据白城地区地形、地质、土壤、植被在一定时间范围内具有相当稳定的特性，选取10月平均水位、汛期6～9月降水量、枯季11～次年3月降水量3个因子，对次年5月平均地下水位进行预测。优化得出的BP网络模型不仅拟合精度高，而且预测效果好。
Based on the stable characteristics of the topography,geology,soil,vegetation within a certain time for Baicheng area,the paper forecasts average groundwater level in May next year by selecting the average water level in October,precipitations from June to September and from November to March.The optimized BP network model has high fitting accuracy and good forecast effect.
出处
《东北水利水电》
2009年第11期32-34,38,共4页
Water Resources & Hydropower of Northeast China
关键词
人工神经网络
BP模型
地下水位
预测
artificial neural network
BP model
groundwater level
forecast
