Research on SSA-BiLSTM-GRU-AM Water Quality Prediction Model

Water quality prediction is of great significance for water circulation and regulation. To enhance the predictive ability of water quality models, a hybrid neural network model optimized by SSA is proposed. The key parameters of the BiLSTM-GRU-AM hybrid network were optimized using the SSA, and an SSABiLSTM-GRU-AM hybrid prediction model was constructed. Using the raw water quality data of a certain sewage treatment plant as the dataset, the Prophet model and Random Forest Algorithm were used to preprocess and screen the features of the raw water quality data, and ablation experiment was set up. In the experiment, the prediction performance of the SSA-BiLSTMGRU-AM hybrid model is superior to other models, with higher prediction accuracy and stronger fitting ability.