Search

Water quality predicted accuracy is beneficial to river ecological management and water pollution prevention. Owing to water quality data has the characteristics of nonlinearity and instability, it is difficult to predict the change of water quality. This paper proposes a hybrid water quality prediction model based on variational mode decomposition optimized by the sparrow search algorithm (SSA-VMD) and bidirectional gated recursive unit (BiGRU). First, the sparrow search algorithm selects fuzzy entropy (FE) as the fitness function to optimize the two parameters of VMD, which improves the adaptability of VMD. Second, SSA-VMD is used to decompose the original data into several components with different center frequencies. Finally, BiGRU is employed to predict each component separately, which significantly improves predicted accuracy. The proposed model is validated using data about dissolved oxygen (DO) and the potential of hydrogen (pH) from the Xiaojinshan Monitoring Station in Qiandao Lake, Hangzhou, China. The experimental results show that the proposed model has superior prediction accuracy and stability when compared with other models, such as EMD-based models and other CEEMDAN-based models. The prediction accuracy of DO can reach 97.8% and pH is 96.1%. Therefore, the proposed model can provide technical support for river water quality protection and pollution prevention. HIGHLIGHTS In order to overcome the problem of nonlinear instability of water quality data, an adaptive decomposition method of improved VMD by SSA is proposed. In the optimization process, fuzzy entropy (FE) is selected as the fitness function to evaluate the complexity between orders.; In order to fully explore the time characteristics of water quality, the BiGRU neural network is used as the prediction model.; The proposed model is applied to the prediction of DO and pH in real rivers.;

Accurate prediction of estuarine salinity can effectively mitigate the adverse effects of saltwater intrusion and help ensure the safety of water resources in estuarine regions. Presently, diverse data-driven models, mainly neural network models, have been employed to predict tidal estuarine salinity and obtained considerable achievements. Due to the nonlinear and nonstationary features of estuarine salinity sequences, this paper proposed a multi-factor salinity prediction model using an enhanced Long Short-Term Memory (LSTM) network. To improve prediction accuracy, input variables of the model were determined through Grey Relational Analysis (GRA) combined with estuarine dynamic analysis, and hyperparameters for the LSTM model were optimized using a multi-strategy Improved Sparrow Search Algorithm (ISSA). The proposed ISSA-LSTM model was applied to predict salinity at the Cangqian and Qibao stations in the Qiantang Estuary of China, based on measured data from 2011–2012. The model performance is evaluated by mean absolute error (MAE), mean absolute percentage error (MAPE), root mean square error (RMSE), and Nash-Sutcliffe efficiency (NSE). The results show that compared to other models including Back Propagation neural network (BP), Gate Recurrent Unit (GRU), and LSTM model, the new model has smaller errors and higher prediction accuracy, with NSE improved by 8–32% and other metrics (MAP, MAPE, RMSE) improved by 15–67%. Meanwhile, compared with LSTM optimized with the original SSA (SSA-LSTM), MAE, MAPE, and RMSE values of the new model decreased by 13–16%, 15–16%, and 11–13%, and NSE value increased by 5–6%, indicating that the ISSA has a better hyperparameter optimization ability than the original SSA. Thus, the model provides a practical solution for the rapid and precise prediction of estuarine salinity.