| 束龙仓,董贵明,陶玉飞,刘丽红.地下河天窗水位变化分析及预测[J].水利学报,2009,40(5): |
| 地下河天窗水位变化分析及预测 |
| Forecast and analysis on water level fluctuation in sinkhole of underground rivers |
| |
| DOI： |
| 中文关键词: 岩溶 地下河天窗 偏最小二乘回归 人工神经网络 水位预测 |
| 英文关键词: karst area sinkhole of underground river partial least squares regression artificial neural network water level forecast |
| 基金项目: |
|
| 摘要点击次数: 3073 |
| 全文下载次数: 1164 |
| 中文摘要: |
| 以贵州省普定县后寨地下河流域内平山天窗水位为研究对象，分析了水位变化规律，并采用偏最小二乘回归、人工神经网络及两种方法的结合建立了高水位阶段水位半日预测模型，对各模型进行了对比分析。结果表明：天窗水位变化划分为枯水位、开采-恢复、高水位3个阶段；高水位阶段天窗水位主要受前半日内的降雨及前半日时的水位影响；偏最小二乘回归是进行天窗水位预测的合适方法。该研究有助于岩溶地区水循环规律的认识及地下河水资源的开发利用。 |
| 英文摘要: |
| The rainfall water level fluctuation of underground rivers' sinkhole in karst area of the Pingshan County, Guizhou Province, China, was analyzed. It is found that the variation of water level can be divided into three phases including the low water level phase, exploitation restoration stage water level phase and high water level phase. It is affected by the earlier stage rainfall within 12 h and corresponding water level. The models for forecasting semidiurnal high water level based on partial least squares regression method, artificial neural network method and the combination of these two methods were established respectively. Comparison of the results of these three models shows that the model based on partial least squares regression method is the best. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |