A novel coupled optimization prediction model for air quality

PM2.5 is a significant pollutant that negatively affects atmospheric environmental sustainability, and accurate prediction of its concentration crucial. Most existing models face challenges such as inadequate data feature capture, dismissal influential factors, subjective model parameter tuning. To address these issues, this paper introduces novel coupled air quality optimization based on Variational Mode Decomposition (VMD), the Informer time series algorithm, Extreme Gradient Boosting (XGBoost), Dung Beetle Optimization Algorithm (DBO). The coupling approach screens features using Spearman coefficient method, optimizes VMD with DBO, decomposes data, classifies various according to approximate entropy. algorithm DBO-optimized XGBoost process different separately, then superimpose reconstruct predicted values obtain results. Using in Nanjing an example, new achieves superior performance (R-squared=0.961, RMSE=1.988, MAE=1.624). Compared WANNs highest accuracy recent relevant studies, our demonstrates 2.96% increase R-squared, 21.89% decrease RMSE, 20.05% MAE. This comparison illustrates proposed DBO-VMD-Informer-XGBoost effectively addresses limitations offers increased accuracy. By employing advanced DBO for innovatively combining VMD, Informer, XGBoost, presents high potential anticipated have broader applications.