地下水NAPLs污染源与含水层参数同步反演识别研究

以往关于地下水NAPLs污染源识别的研究中常假设含水层参数为已知，而在实际问题中，某些关键含水层参数分布常难以准确获取，针对此问题，本项目对地下水NAPLs污染源与含水层参数进行同步反演研究。研究应用组合替代模型表征地下水NAPLs污染源、含水层参数与污染物浓度响应关系，以提高同步反演地下水NAPLs污染源与含水层参数识别结果的准确性与计算效率。然后，构建基于双层规划的地下水NAPLs污染源与含水层参数同步识别优化模型，并探讨优化模型的准确求解方法，进一步提高反演结果的准确性。最后，通过多组物理模型实验结合数值实验，验证反演结果的有效性，为应用其解决实际问题提供理论基础。研究成果可为地下水污染肇事者责任的判定、地下水污染修复方案的设计、地下水风险评价等工作提供决策依据。

It is common to assume the aquifer parameters are known in previous groundwater pollution source identification studies. However, some key aquifer parameters are hard to gain in practice problems. Thus, this program focuses on simultaneous identification aquifer parameters and groundwater pollution sources. To improve the accuracy and efficiency of this problem, ensemble surrogate model is explored to reflect the relationship between groundwater pollution source, aquifer parameters and contaminant concentrations. To improve the accuracy of inverse results, bi-level programming is discussed. To make clear the effective solution of simultaneous identification aquifer parameters and groundwater pollution sources, cooperative strategy and quantum-behaved particle swarm optimization is combined to study. The above inverse methods are tested by multiple physical model experiments combined with numerical experiments. Research results can help to judge the pollution liability, design the groundwater pollution remediation plan and risk assessment.