某核电厂水文地质参数的遗传反演  

Genetic algorithm inversion for hydrogeological parameters in a nuclear power plant

作　　者：王小明[1] 罗长保[1] 任国澄[1] 

机构地区：[1]中国能源建设集团广东省电力设计研究院,广州510663

出　　处：《工程勘察》2013年第4期39-43,共5页Geotechnical Investigation & Surveying

基　　金：中国博士后科学基金(2011M500187);广东省自然科学基金(S2011040005245);广东省电力设计研究院科标项目(EV00711W)

摘　　要：利用非稳定流抽水试验资料,以Theis公式为计算分析模型,采用遗传算法思想,以渗透系数K和贮水系数μ*为搜索变量对模型进行优化,得到一种在全局范围内进行随机搜索,并能自适应地控制搜索过程求得最优解的方法。最后结合某核电工程的计算实例,验证了该方法的合理性和可靠性,并与配线法计算结果进行了简要对比,显示出遗传算法的优点。By using the results from the unsteady flow pumping tests,taking Theis formula as the analysis model and optimizing the model with the permeability coefficient K and the storativity factor μ＊ as the search variables based on the genetic algorithm,a kind of method that can search the optimal parameters in the global range randomly and control the searching process adaptively is obtained.The engineering application case in a nuclear power plant indicates that the method is rational and reliable,and its merit is shown by comparing with the results from the curve fitting method.

关 键 词：抽水试验 水文地质参数 遗传算法 配线法 反演 

分 类 号：P641.2[天文地球—地质矿产勘探]