Groundwater Level Prediction for the Arid Oasis of Northwest China Based on the Artificial Bee Colony Algorithm and a Back-propagation Neural Network with Double Hidden Layers

摘要: Groundwater is crucial for economic and agricultural development, particularly in arid areas where surface water resources are extremely scarce. The prediction of groundwater levels essential understanding dynamics providing scientific guidance the rational utilization resources. A back propagation (BP) neural network based on artificial bee colony (ABC) optimization algorithm was established this study to accurately predict overexploited Northwest China. Recharge, exploitation, rainfall, evaporation were used as input factors, whereas level output factor. Results showed that fitting accuracy, convergence rate, stabilization ABC-BP model better than those particle swarm (PSO-BP), genetic (GA-BP), BP models, thereby proving can be a new method predicting levels. with double hidden layers topology structure 4-7-3-1, which overcame overfitting problem, developed Yaoba Oasis from 2019 2030. results different mining regimes area will gradually decrease exploitation quantity increases then undergo decline stage given existing condition 40 million m3/year. According simulation under scenarios, most appropriate amount should maintained at 31 m3/year promote sustainable development Oasis.