Surrogate Model-Based Simulation-Optimization Approach for Groundwater Source Identification Problems

This study investigates and discusses a time-efficient technology that contains a surrogate model within a simulation-optimization model to identify the characteristics of groundwater pollutant sources. In the proposed surrogate model, Latin hypercube sampling (a stratified sampling approach) and artificial neural network (commencing at the stress period when the concentration is within a certain range, and ending at the peak time) were utilized to reduce workload and costly computing time. The results of a comparison between the proposed surrogate model and the common artificial neural network model and non-surrogate model indicated that the proposed model is a time-efficient technology which could be used to solve groundwater source identification problems.     

Optimization-Based Atmospheric Plume Source Identification

This work applies optimization and an Eulerian inversion approach presented by Bagtzoglou and Baun in 2005 in order to reconstruct contaminant plume time histories and to identify the likely source of atmospheric contamination using data from a real test site for the first time. Present-day distribution of an atmospheric contaminant plume as well as data points reflecting the plume history allow the reconstruction and provide the plume velocity, distribution, and probable source. The method was tested to a hypothetical case and with data from the Forest Atmosphere Transfer and Storage (FACTS) experiment in the Duke experimental forest site. In the scenarios presented herein, as well as in numerous cases tested for verification purposes, the model conserved mass, successfully located the peak of the plume, and managed to capture the motion of the plume well but underestimated the contaminant peak.     

Breakthrough Curves Characterization and Identification of an Unknown Pollution Source in Groundwater System Using an Artificial Neural Network (ANN)

Contamination of groundwater constrains its uses and poses a serious threat to the environment. Once groundwater is contaminated, the cleanup may be difficult and expensive. Identification of unknown pollution sources is the first step toward adopting any remediation strategy. The proposed methodology exploits the capability of a universal function approximation by a feed-forward multilayer artificial neural network (ANN) to identify the sources in terms of its location, magnitudes, and duration of activity. The back-propagation algorithm is utilized for training the ANN to identify the source characteristics based on simulated concentration data at specified observation locations in the aquifer. Uniform random generation and the Latin hypercube sampling method of random generation are used to generate temporal varying source fluxes. These source fluxes are used in groundwater flow and the transport simulation model to generate necessary data for the ANN model-building processes. Breakthrough curves obtained for the specified pollution scenario are characterized by different methods. The characterized breakthrough curves parameters serve as inputs to ANN model. Unknown pollution source characteristics are outputs for ANN model. Experimentation is also performed with different number of training and testing patterns. In addition, the effects of measurement errors in concentration measurements values are used to show the robustness of ANN based methodology for source identification in case of erroneous data.     

等效排气筒有效高度计算方法的比选及其应用于大气污染预测的可行性

目前,等效排气筒多用于大气污染物总量控制,其预测精度和范围的不明晰限制了其在污染物运移扩散领域的进一步应用。基于《大气污染物综合排放标准》,依据所预测的范围和浓度精度对8种典型等效计算方法进行了比选,并验证了将等效排气筒用于不同工况下污染物运移扩散预测的可行性。改进的有效高度等效算法 (源强加权算术平均法) 综合考虑了不同高度和源强参数特征,以2个排放同种污染物的相邻排气筒为例,所计算的高斯模式下等效后下风向污染物浓度场总体分布趋势与等效前叠加计算结果一致,且预测精度优于《大气污染物综合排放标准》中提出的均方根平均法和其他等效算法。对不同风速条件下 (1.5~4.5 m·s⁻¹) 等效前后下风向污染物浓度场分布计算比较,发现即使风速改变仍可保证较高的最大落地浓度预测精度 (−6.87%~−2.21%),特别是风速较大时其预测精度更高 (达到−2.21%) 。这验证了该方法的有效性和稳定性。本研究探讨的源强加权算术平均值算法,进一步提升了等效排气筒相关参数计算的合理性,并拓展了其在大气预测评价中的应用。