引用本文:
【打印本页】   【下载PDF全文】   查看/发表评论  下载PDF阅读器  关闭
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 2397次   下载 1796 本文二维码信息
码上扫一扫!
分享到: 微信 更多
土壤污染耦合模型的有限元模拟及参数反演
郝竹林1, 闵 涛2, 谷明礼2
1.西安理工大学 水利水电学院;2.西安理工大学 理学院
摘要:
【目的】对污染物在土壤环境中的运移分布规律进行模拟,并对模型中的主要参数进行反演。【方法】以土壤中的硝酸盐氮和氨氮为研究对象,建立一维数学耦合模型,用建立的模型对不同时间和不同深度土壤的硝酸盐氮和氨氮含量变化进行分析,并给出其求解的有限元过程。在此基础上,采用微分进化算法对建立的一维耦合模型的多个参数进行反演。【结果】构建了土壤污染一维耦合模型,用该模型对不同时间和不同深度土壤的硝酸盐氮和氨氮含量变化进行分析,结果表明,污染物在土壤中的含量达到一定值后会趋于稳定或降低,说明污染物在土壤中先有一个积累过程,然后有一个释放过程。采用微分进化算法反演得到的一维耦合模型的主要参数值与真值较为接近。【结论】构建的一维耦合模型可以定量地研究污染物在土壤中的运移分布规律;遗传进化反演算法可为土壤水分运动及溶质运移数值模拟提供所需的基本参数。
关键词:  土壤污染  有限元方法  微分进化  参数反演
DOI:
分类号:
基金项目:国家自然科学基金项目(50979088)
The numerical solution and parameter inversion in the couple model of soil pollution
Abstract:
【Objective】The research was conducted to study the law of pollutants in the soil environment migration distribution,and inverse the main parameters of the model.【Method】With the soil of ammonia nitrogen and nitrate nitrogen as the research object,a mathematical coupled model was established and the solution of the finite element process was given.On this basis,this coupled multiple parameters of the model were inverted by differential evolution algorithm.【Result】For different time and depths of soil,contents of nitrate nitrogen and ammonia nitrogen transformation were analyzed by building One-dimensional coupled model.The result indicates that the content of pollutants in the soil will stabilize or lower after reaching a certain value,which firstly has a cumulative process of pollutants in the soil and then there is a release process.Using differential evolution algorithm for the main parameters of one-dimensional coupled model,the inversion value is close to the true value.【Conclusion】Established One-dimensional coupled model can be used to study the law of the transport and distribution of pollutants in the soil.The differential evolution algorithm provides the necessary basic parameters for soil water movement and solute transport numerical simulation.
Key words:  soil pollution  finite element method  differential evolution algorithm  parameter inversion