Simulation of Pb2+ Adsorption in Porous Media Based on Different BP Neural Networks

In order to study the adsorption process model of Pb
²⁺
containing wastewater by porous media, based on genetic algorithm, particle swarm optimization and BP neural network algorithm, 17 groups of real experimental data were used as training samples, and three adsorption models were established with adsorbent dosage, adsorption time and the value of pH as input variables and initial concentration of Pb
²⁺
as output variable. Compared with the GA-BP neural network model and the PSO-BP neural network model, the prediction results of the BP neural network model output are in good agreement with the experimental data. The Pearson correlation coefficient between the BP prediction value and the original data is 0.991, which shows that the model is feasible and effective for the simulation of the adsorption process of Pb
²⁺
in porous media. The construction of three BP neural network models will greatly enrich the simulation of various chemical adsorption experiments and has broad development prospects.