Article
Engineering, Civil
Xue Jiang, Rui Ma, Yanxin Wang, Wenlong Gu, Wenxi Lu, Jin Na
Summary: This study proposes a new two-stage surrogate-assisted Markov chain Monte Carlo-based Bayesian framework for identifying contaminant source parameters in groundwater. An adaptive update feedback process and a multiobjective feasibility-enhanced particle swarm optimization algorithm are utilized to enhance the accuracy and efficiency of the framework.
JOURNAL OF HYDROLOGY
(2021)
Article
Engineering, Civil
Zibo Wang, Wenxi Lu, Zhenbo Chang, Jiannan Luo
Summary: In this paper, a improved butterfly optimization algorithm is proposed and combined with Ensemble Kalman filter and optimization methods to build a more robust and practical combined search method (CSM) for groundwater contamination source identification. The CSM significantly improves the identification accuracy and effectiveness compared with any single method.
JOURNAL OF HYDROLOGY
(2023)
Article
Engineering, Multidisciplinary
Jiannan Luo, Yong Liu, Xueli Li, Xin Xin, Wenxi Lu
Summary: In this study, a two-stage adaptive surrogate model-assisted trust region GA (TSASM-TRGA) framework was developed to improve the accuracy and stability of groundwater contamination source inversion. The results showed that the TSASM-TRGA framework outperformed other frameworks in terms of accuracy and stability.
APPLIED MATHEMATICAL MODELLING
(2022)
Article
Computer Science, Interdisciplinary Applications
Jiexiang Hu, Lili Zhang, Quan Lin, Meng Cheng, Qi Zhou, Huaping Liu
Summary: The paper proposed a new multi-fidelity surrogate model-based robust optimization method to address prediction uncertainty, achieving better optimal design solutions.
STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION
(2021)
Article
Mathematics
Zebin Zhang, Martin Buisson, Pascal Ferrand, Manuel Henner
Summary: The cokriging method can enhance the accuracy of the surrogate model and reduce modeling time, but the high computational cost of high order derivatives poses a bottleneck for derivative enhanced methods.
Article
Computer Science, Interdisciplinary Applications
Aatish Anshuman, T. I. Eldho
Summary: Contaminants in groundwater may come from different sources that need to be identified for informed decision-making in remediation. In the early stages of aquifer contamination, the sources are generally unknown. Solving the governing equations of contaminant transport backward in time to estimate unknown release histories is an ill-posed inverse problem. Observation errors can induce uncertainty in breakthrough curves. A simulation-optimization (SO) model based on meshless radial point collocation method and multiverse optimizer (MVO) is proposed in this study. The proposed model outperforms two other SO models in terms of release history estimation in groundwater.
JOURNAL OF HYDROINFORMATICS
(2023)
Article
Engineering, Civil
Majid Vali, Mohammad Zare, Saman Razavi
Summary: This study introduces a novel 'local surrogate modelling' framework guided by automatic clustering for solving computationally intensive groundwater remediation optimization problems. Results show that the proposed automatic clustering-based local surrogate modeling is effective and reliable while reducing at least 60 percent of the computational burden.
JOURNAL OF HYDROLOGY
(2021)
Article
Engineering, Civil
Zheng Han, Wenxi Lu, Yue Fan, Jianan Xu, Jin Lin
Summary: A new stochastic S/O framework is proposed and applied to a real-world case in China, overcoming limitations of traditional multi-objective evolutionary algorithms by introducing information entropy theory. Additionally, a surrogate model using MGGP method is developed, significantly reducing computational burden.
WATER RESOURCES MANAGEMENT
(2021)
Article
Environmental Sciences
Jiuhui Li, Zhengfang Wu, Hongshi He, Wenxi Lu
Summary: The location and release history of groundwater contaminant sources (GCSs) are usually unknown, but previous studies have used prior information to identify GCSs. This study compared two methods for GCSs identification and found that the S/O method is more suitable for GCSs identification than the EnKF method.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2022)
Article
Acoustics
Yu Bai, Zhenrui Peng, Zenghui Wang
Summary: This study proposes a finite element (FE) model updating method based on a trust region (TR) and an adaptive surrogate model. The method uses sampling to realize the inverse identification of model parameters. It introduces an average sensitivity method to select the parameters to be updated and converts the FE model updating problem into a surrogate model optimization problem. Through adaptive sampling and the use of the TR method, the method achieves high updated accuracy and efficiency.
JOURNAL OF SOUND AND VIBRATION
(2023)
Article
Engineering, Civil
Mengtian Wu, Lingling Wang, Jin Xu, Zhe Wang, Pengjie Hu, Hongwu Tang
Summary: This paper proposes a multi-objective ensemble surrogate-based optimization algorithm named MESOA for groundwater optimization designs. By using surrogate models with various basis functions and an adaptive switching technique, MESOA can fully depict the outline of the true Pareto front with limited simulation invocations.
JOURNAL OF HYDROLOGY
(2022)
Article
Engineering, Multidisciplinary
Jie Qu, Xiao-Yao Han
Summary: This article introduces an adaptive multi-surrogate constrained optimization method (AMSCOM) that automatically determines the appropriate metamodel for each black-box function in the constrained optimization problem (COP) and finds the optimum concurrently.
INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING
(2021)
Article
Environmental Sciences
Shuangsheng Zhang, Jing Qiang, Hanhu Liu, Xueqiang Zhu, Hongli Lv
Summary: This paper proposed a conservative adaptive Kriging surrogate model to address the problems associated with using surrogate models, such as large training sample size, low accuracy, and poor optimization results. By coupling the Kriging surrogate model, optimal solution adaptive sampling method, and conservative prediction idea, the CAKSM effectively constrained pollutant mass concentrations within a controlled value, improving the reliability of the optimization scheme.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2022)
Article
Water Resources
F. Poursalehi, A. Akbarpour, S. R. Hashemi
Summary: In this research, a simulation-optimization model using Isogeometric analysis and the Grey wolf optimization algorithm was proposed to determine the optimal location of injection wells. The results show that the model has high accuracy and after constructing injection wells in the optimal location, the groundwater table rises significantly. The study also highlights the importance of hydraulic conductivity and specific yield in determining the optimal location.
APPLIED WATER SCIENCE
(2022)
Article
Engineering, Environmental
Jiuhui Li, Zhengfang Wu, Hongshi He, Wenxi Lu
Summary: The grey wolf optimization algorithm (GWO) has the disadvantage of premature convergence when solving the optimization model for groundwater contamination sources (IGCSs) due to its weak local search ability. To improve it, a hybrid grey wolf gradient optimization algorithm (HGWGO) was developed by integrating GWO with the gradient descent algorithm, which showed a strong local search ability and less dependence on the initial value. The HGWGO was applied to the optimization model to enhance the accuracy of IGCSs. Additionally, a surrogate model using a deep belief neural network (DBNN) was established to participate in the iterative calculation, reducing the computational load and time consumption.
STOCHASTIC ENVIRONMENTAL RESEARCH AND RISK ASSESSMENT
(2023)