Search

Groundwater contamination source identification (GCSI) is critical for taking effective measures to protect groundwater resources, assess risks, mitigate disasters, and design remediation strategies. Simulation–optimization techniques have been effective tools for GCSI. However, previous studies have applied individual surrogate models when replacing simulation models, rather than making efforts to combine various methods to improve the approximation accuracy of the surrogate model over the simulation model. In this study, the kernel extreme learning machine (KELM) model was proposed to enhance the surrogate model, and to approach GCSI problems, especially those of dense nonaqueous phase liquid-contaminated aquifers, more effectively. In addition, a kriging model and a support vector regression (SVR) model were built and compared with the KELM model, and various ensemble surrogate (ES) modeling techniques were applied to establish four ES models. Results showed that the KELM model was more accurate than the kriging and SVR models; however, the ES models performed much better than the three individual surrogate models. The most precise ES model increased the certainty coefficient (R2) to 0.9837, whereas limiting the maximum relative error to 13.14%. Finally, a mixed-integer nonlinear programming optimization model was established to identify the groundwater contamination source in terms of location and release history, and simultaneously assess aquifer parameters. The ES model developed in this article could reasonably predict the system response under given operation conditions. Replacement of the simulation model by the ES model considerably reduced the computation burden of the simulation–optimization process and simultaneously achieved high computation accuracy. [ABSTRACT FROM AUTHOR]

Surfactant-enhanced aquifer remediation (SEAR) is an efficient way for clearing dense nonaqueous phase liquids (DNAPLs) which may result in serious environment and health threats. To limit the high cost of SEAR, simulation optimization techniques are generally applied to ensure that an optimal remediation strategy is achieved. Furthermore, surrogate model techniques have been widely used to reduce the high computational burden associated with these processes. However, previous research rarely involved comparison of different surrogate models for multiphase flow numerical simulation models. In this regard, we conducted a comparative analysis to select the optimal modeling technique and parameter optimization algorithm for surrogate models in DNAPL-contaminated aquifer remediation strategy optimization problems. Latin hypercube sampling method was used to collect data in the feasible region of input variables. Surrogate models were developed using radial basis function artificial neural network, Kriging, and support vector regression. Genetic algorithm, self-adaptive particle swarm optimization (PSO), and self-adaptive PSO based on simulated annealing (SA) were applied to optimize the parameters of the surrogate model. Results showed that the optimal surrogate model was Kriging model with the parameters obtained by self-adaptive PSO based on SA. Relative errors of the contaminant removal rates between the optimal surrogate model and simulation model for 100 validation samples were lower than 3.5%, clearly confirming the optimal performance of the proposed model. Finally, computation of run time enabled us to conclude that the surrogate model presented in this article was capable of considerably reducing computational burden of simulation optimization processes. [ABSTRACT FROM AUTHOR]