Department of Civil Engineering, Faculty of Engineering, Architecture and Science, Ryerson University, 350 Victoria Street, Toronto, Ontario, Canada M5B 2K3.
J Hazard Mater. 2010 Apr 15;176(1-3):521-6. doi: 10.1016/j.jhazmat.2009.11.060. Epub 2009 Nov 14.
Solving groundwater remediation optimization problems based on proxy simulators can usually yield optimal solutions differing from the "true" ones of the problem. This study presents a new stochastic optimization model under modeling uncertainty and parameter certainty (SOMUM) and the associated solution method for simultaneously addressing modeling uncertainty associated with simulator residuals and optimizing groundwater remediation processes. This is a new attempt different from the previous modeling efforts. The previous ones focused on addressing uncertainty in physical parameters (i.e. soil porosity) while this one aims to deal with uncertainty in mathematical simulator (arising from model residuals). Compared to the existing modeling approaches (i.e. only parameter uncertainty is considered), the model has the advantages of providing mean-variance analysis for contaminant concentrations, mitigating the effects of modeling uncertainties on optimal remediation strategies, offering confidence level of optimal remediation strategies to system designers, and reducing computational cost in optimization processes.
基于代理模拟器解决地下水修复优化问题通常可以产生不同于问题“真实”解的最优解。本研究提出了一种新的随机优化模型，即建模不确定性和参数确定性下的随机优化模型（SOMUM），以及相关的求解方法，用于同时解决与模拟器残差相关的建模不确定性和优化地下水修复过程。这是与之前建模工作的一个新的尝试。之前的工作侧重于解决物理参数（即土壤孔隙率）的不确定性，而这一次则旨在解决数学模拟器（由模型残差引起）的不确定性。与现有的建模方法（即只考虑参数不确定性）相比，该模型具有提供污染物浓度的均值-方差分析、减轻建模不确定性对最优修复策略的影响、为系统设计者提供最优修复策略的置信水平以及降低优化过程计算成本的优点。
