Inversion Analysis of Impervious Curtain Permeability Coefficient Using Calcium Leaching Model, Extreme Learning Machine, and Optimization Algorithms

Abstract

The calcium leaching effect leads to a decrease in the impermeability of the impervious curtain. The inverse analysis strategy was introduced in this study because the calcium leaching parameters of the curtain are not easy to determine. An orthogonal design and the finite element method were used in the strategy. The time-series data of hydraulic head and leakage volume were applied to construct the objective function. The extreme learning machine (ELM) was proposed to build the reflection sets. Genetic algorithm (GA), simulated annealing (SA), sparrow search algorithm (SSA), and particle swarm optimization (PSO) were employed to accelerate the iterative search for the target parameters. The target parameters of the calcium leaching model were used for finite element verification by comparing the monitored and simulated values. The simulated values of hydraulic head and leakage by PSO show good agreement with measurements. The evolution of the curtain permeability coefficient in 100 years was analyzed. The results demonstrate the strategy’s feasibility in determining the curtain’s calcium leaching parameters and permeability coefficients.