基于遗传算法的地下水人工回灌系统优化——以北京永定河冲洪积扇为例  

Optimization of artificial recharge of groundwater system based on parallel genetic algorithm——a case study in the alluvial fan of Yongding River in Beijing

作　　者：郝奇琛[1] 邵景力[1] 李宇[2] 谢振华[2] 崔亚莉[1] 

机构地区：[1]中国地质大学,北京100083 [2]北京市水文地质工程地质大队,北京100195

出　　处：《南水北调与水利科技》2015年第1期67-71,共5页South-to-North Water Transfers and Water Science & Technology

基　　金：国土资源部公益性行业科研专项经费资助(201211079-4);国家重点研究基础发展计划"973"项目(2010CB428804)

摘　　要：地下水人工回灌是增加含水层补给资源量的有效方法之一。为评价地下水人工回灌系统的最大回灌效率。以地下水长期超采的北京永定河冲洪积扇为例,建立了地下水人工回灌系统优化模型。为预测地下水对回灌方案的响应,优化模型中嵌入了地下水流模型,模拟预测的结果返回到优化模型中作为约束条件,优化过程最终由遗传算法(GA)求解。GA的选择方法为轮盘赌选择法,并使用罚函数法解决了优化模型的限制条件问题。求解结果显示,优化后的地下水人工回灌系统将更有效的增加含水层的储存资源量,在地下水限高水位约束下,两种方案回灌至含水层的水量将从127.42×106 m3和243.48×106 m3分别增加到140.46×106 m3和275.55×106 m3,回灌结束后水位最大升幅为26m。说明开发的优化模型可以高效求得最优回灌方案。Artificial recharge is one of the most efficient approaches to increase the aquifer groundwater storage.In order to evaluate the maximum recharge rate of the artificial recharge system to groundwater,an optimization model of artificial recharge system was developed for the alluvial fan in the Yongding river of Beijing,where groundwater is often over extracted.Meanwhile,agroundwater flow model was integrated into the optimization model to predict the response of groundwater to the artificial scheme.The groundwater model results were used as the constraint conditions in the optimization model,which was solved by the genetic algorithm(GA).Roulette wheel selection was used for GA,and penalty function method was used to solve the constraints of the optimization problem.The optimal results indicated that the optimal artificial recharge system can increase the aquifer groundwater storage effectively.The aquifer storage can increase from 127.42×106 m3 and 243.48×106 m3 to 140.46×106 m3 and 275.55×106 m3 under two recharge schemes without exceeding the upper limit of groundwater level,and groundwater level can increase 26 mafter the optimization.The proposed optimization model is able to determine the optimal recharge scheme effectively.

关 键 词：优化模型 遗传算法 地下水人工补给 回灌速率 地下水流模型 

分 类 号：P641.25[天文地球—地质矿产勘探]