Two-step approach based multi-objective groundwater remediation in highly heterogeneous media using enhanced random vector functional link integrated with evolutionary marine predator algorithm

Abstract

We herein propose two-step approach based simulation-optimization models for groundwater remediation using enhanced random vector functional link (ERVFL) and evolutionary marine predator algorithm (EMPA). The weighted least square method is used to improve the robustness of the ERVFL network, where weights are computed using the kernel density estimator (KDE). The EMPA is developed by modifying the marine predator algorithm (MPA) using elite opposition-based learning, biological evolution operators, and elimination mechanisms. In the multi-objective version of EMPA, the non-dominated solutions are stored in an external repository using an archive controller and adaptive grid mechanism to promote better convergence and diversity of the Pareto front. The performance evaluation of EMPA on several test functions suggests its superiority over other metaheuristics for both single-objective and multi-objective optimization. The ERVFL network is then used to approximate the finite difference based groundwater flow and transport models to accelerate computational efficiency. The two-step approach based S-O models are then developed by integrating the simulation models directly or through the ERVFL network with the EMPA. The first step aims to find optimal pumping locations using EMPA with combinatorial optimization technique by minimizing the percentage of contaminant mass remained in the aquifer. In the second step, the ERVL based proxy simulator is coupled with EMPA and used for multi-objective optimization while explicitly using the pumping well locations as obtained in the first step. The multi-objective optimization generates a Pareto-optimal solution representing the relationship between the water extraction rates and the amount of contaminant mass in the aquifer. Further analyses suggest that the two-step approach shows a significant advantage over the traditional methods for multi-objective groundwater remediation.