Dissolved Oxygen Prediction Based on Long short-Term Memory Network Optimized by Improved Shuffled Frog-Leaping Algorithm

For the purpose of predicting the level of dissolved oxygen during the sewage treatment process, this paper is based on the improved Shuffled Frog-Leaping algorithm, establishes a long-short-term memory neural network model to realize the prediction and control of dissolved oxygen concentration in the process of sewage treatment. The Tent chaotic mapping strategy is introduced into MSFLA, a mutation strategy is added to the global worst frog individual position and an adaptive disturbance strategy is added to the individual. Finally, MSFLA is used to optimize the parameters of LSTM network, and the MSFLA-LSTM sewage treatment dissolved oxygen prediction model is established. Combined with the actual data of a sewage plant for prediction analysis, and set up multiple groups of comparative experiments, the results show that the MSFLA-LSTM dissolved oxygen prediction model is more accurate and stable.