联合EEMD与BP神经网络的灌区水源情势预测研究

吴小宏; 刘招; 王丽霞; 李强; 李雯晴

引用本文:	吴小宏,刘招,王丽霞,等.联合EEMD与BP神经网络的灌区水源情势预测研究[J].灌溉排水学报,0,():-.
	WU Xiaohong,LIU Zhao,WANG Lixia,et al.联合EEMD与BP神经网络的灌区水源情势预测研究[J].灌溉排水学报,0,():-.

【打印本页】【下载PDF全文】【查看/发表评论】【EndNote】【RefMan】【BibTex】

过刊浏览高级检索

本文已被：浏览 1534次下载 0次
分享到：微信更多字体:加大+\|默认\|缩小-
联合EEMD与BP神经网络的灌区水源情势预测研究
吴小宏¹, 刘招², 王丽霞³, 李强⁴, 李雯晴⁴
1.泾惠渠灌溉管理局;2.长安大学水与发展研究院;3.长安大学地质工程与测绘学院;4.长安大学旱区地下水文与生态效应教育部重点实验室

摘要:

泾河作为泾惠渠灌区渠首地表水源，来水逐年减少，【目的】本文研究为了分析探讨泾惠渠灌区渠首水源形势，保障灌区水资源管理及粮食生产安全。【方法】研究建立基于EEMD-BP的水文序列预测模型，通过对模型进行训练和校正，最终预测了未来十年泾河来水来沙形势，分析了渠首的水资源可利用量。【结果】研究结果表明：将EEMD与BP神经网络二者结合，可有效发挥各自优势，大大增强对非平稳时间序列的预测能力，验证期各项评估指标也较为理想；灌区渠首水沙变化较为同步，在未来一定时期二者趋势均为短暂的上升后再下降，年均径流量约为11.86亿m3，较当前略有上升，年均输沙量约为1亿t，较当前略有下降，预测结果延续了泾河水沙多年变化的大致趋势；灌区渠首可利用水资源量相对比较平稳，平均水资源可利用量约为7.79亿m3，可满足引水灌溉需求，但应注意一些干旱年份水资源的调蓄利用，谨防发生灌溉水短缺问题。【结论】验证表明模型误差在可接受范围内，且预测系列的变化趋势与当前实测系列完全一致，建立的EEMD-BP模型预测年径流及泥沙效果良好，预测结果具有较高的可信度。

关键词: 水资源;年径流量;预测;水沙规律;泾惠渠灌区

DOI：

分类号:S273

基金项目:

The Surface Water Resources Situation and Its Forecasting for Jinghuiqu Irrigation Area Based on EEMD-BP

WU Xiaohong¹, LIU Zhao², WANG Lixia³, LI Qiang⁴, LI Wenqing⁴

1.The Administration Bureau of Jinghui Canal Irrigation Area,Sanyuan;2.Institute of Water and development,Chang’an University,Xi’an;3.College of Geological Engineering and Geomatics,Chang’an University,Xi’an;4.Key Laboratory of Subsurface Hydrology and Ecological Effects in Arid Region of Ministry of Education,Chang’an University,Xi’an

Abstract:

【Objective】To probe the problem of water resources reduction and its trend at Zhangjiashan, where the head of Jinghui Irrigation Area lies, 【Method】a hydrological series prediction model based on EEMD and BP artificial neural network was established, by training and correcting the model, the runoff and sediment situation at Zhangjiashan in the next ten years is predicted, and the available surface water resources for the irrigation area are analyzed. 【Result】The results show that the combination of EEMD and BP neural network can give full play to their respective advantages and greatly enhance the prediction ability for non-stationary time series. The change of runoff and sediment at the irrigation area's canal head is relatively synchronous, rise in the near future and then decline. The average annual runoff is about 1.186 billion m3, slightly higher than that at present, and the average annual sediment transport is about 100 million t, slightly lower than that at present. 【Conclusion】The available water resources at Zhangjiashan is relatively stable, the average is about 779 million m3, which can meet the surface water demand of the irrigation area. However, attention should be paid to some dry years to prevent the shortage of irrigation water.

Key words: water resources; annual runoffs; forecasting; nature of runoff and sediment; Jinghuiqu Irrigation Area