Prediction technology for traffic capacity in congested water areas of ships
Aiming at the problem of high difficulty in predicting traffic capacity in congested water areas due to high ship traffic flow and busy waters,a technology for predicting traffic capacity in congested water areas is proposed.Determ-ine the ship's status and establish a ship navigation queue model by utilizing the ship's coordinate position and navigation speed.Based on the Fujino ship domain model,determine the elliptical area corresponding to ship navigation as the domain of ship congestion waters.Determine the relevant factors that affect the capacity of shipping channels within the field of ships.Set indicators such as ship density,ship speed,and other factors that affect water traffic capacity as inputs for the long short-term memory network.The long short-term memory network uses three gating units,namely the forget gate,input gate,and output gate,to select whether to retain or forget the input data,and output the prediction results of ship traffic capacity in congested water areas.The experimental results show that this technology can effectively predict the traffic capacity of ship congested waters,improve the utilization rate of waterways,and ensure the safety of water traffic.
ship congested waterstraffic capacity predictionship fieldforgetting gategate control unit
NETL
NSTL
万方数据
