Abstract: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.

Ship Traffic Flow Forecast of Qingdao Port Based on LSTM

K-Nearest Neighbor Model Based Short-Term Traffic Flow Prediction Method

A Survey of Traffic Flow Prediction Methods Based on Long Short-Term Memory Networks

A Deep Learning Model for Ship Trajectory Prediction Using Automatic Identification System (AIS) Data.

Research on Short-Term Prediction of Typical Trial Sea Environment in China Based on LSTM Neural Network

Ship Traffic Flow Prediction in Wind Farms Water Area Based on Spatiotemporal Dependence

Real-Time Prediction of Multi-Degree-of-Freedom Ship Motion and Resting Periods Using LSTM Networks

QSD-LSTM: Vessel trajectory prediction using long short-term memory with quaternion ship domain

Prediction Model of Ship Arrival Time using Neural Network and Kalman Filter

Ship Navigation Behavior Prediction based on AIS Data

Salinity Prediction Based on Improved LSTM Model in the Qiantang Estuary, China

Short-Term Prediction in Vessel Heave Motion Based on Improved LSTM Model

A novel MP-LSTM method for ship trajectory prediction based on AIS data

Traffic Flow Combination Forecasting Method Based on Improved LSTM and ARIMA

Multiscale attention-based LSTM for ship motion prediction

Artificial Intelligence in Ship Trajectory Prediction

A New Method of Inland Water Ship Trajectory Prediction Based on Long Short-Term Memory Network Optimized by Genetic Algorithm

Short-Term Passenger Flow Prediction of Railway Station Based on Improved LSTM

An AIS Data-Driven Dynamic Spatiotemporal Prediction Model for Ship Traffic Flow in the Port

Ship Fuel Consumption Prediction Model Based on LSTM

A Ship Trajectory Prediction Framework Based on a Recurrent Neural Network