Decomposition prediction and optimal ensemble strategy improve river dissolved oxygen prediction accuracy

Abstract

The accurate prediction of dissolved oxygen (DO) concentration in rivers is very important for the management of aquatic ecosystems, However, the hybrid model of ' modal decomposition + prediction ' for predicting the nonlinear change of dissolved oxygen in rivers is still insufficient. In this paper, a frequency division prediction framework based on the optimal ensemble of Complementary Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) is proposed. The dissolved oxygen sequence was decomposed into multiple components by CEEMDAN, and the long short-term memory network (LSTM), support vector regression (SVR) and multi-layer perceptron (MLP) models were constructed to predict each component independently. An innovative grid search algorithm with constraints is constructed, and the advantages of each model are complemented by dynamic combination. The optimal ensemble scheme is obtained with the goal of minimizing the mean absolute error (MAE) of the training set. The empirical study of monitoring sections A and B in the Ganjiang River Basin shows that : in the prediction task, the prediction of the training set, the MAE of the integrated model is 18.6–35.5% lower than that of the ensemble model, the root mean square error (RMSE) is 22.1–22.8% lower, and the determination coefficient (R2) reaches 0.954 and 0.972. In particular, the error accumulation of MAE in the 3-day prediction is 27.2–81.4% lower than that of the mixed model. This framework enables the modes of multi-component dissolved oxygen series prediction to be effectively aliasing, and provides an extensible technical path for the intelligent management of the basin.