Navigation Path Optimization Of Merchant Ship And Fishing Ship Based On Traffic Flow Prediction

Chengshantou waters,located at the easternmost end of Shandong Peninsula,is a traffic fortress for ships entering and leaving Bohai Bay and ports in the northern Yellow Sea.At the same time,Chengshantou water area is also an important fishery industry area in northern China.The overlapping of merchant ship navigation and fishery operation areas leads to mixed navigation of merchant ships and fishing boats in Chengshantou water area,which is prone to ship accidents.Chengshantou water area has become a high incidence area of collision between merchant ships and fishing boats in China.In this paper,deep learning is used to study the optimization of ship traffic flow and navigation path,focusing on the problem of collision of commercial fishing ships and combining with the environmental characteristics of the actual navigable waters in Chengshantou.Firstly,the basic situation of natural environment and navigation environment of Chengshantou water area is analyzed,and the ship accidents in this water area from 2016 to 2019 are counted,and the causes of collision accidents of commercial fishing ships in the water area are analyzed from four aspects: merchant ships,fishing ships,environmental factors and management system by using fault tree method.Then,the data of Automatic Identification System(AIS)in Chengshantou traffic separation waters are preprocessed and the data of ship traffic flow are mined.In order to accurately predict the ship traffic flow in Chengshantou traffic separation waters,Bi-GRU neural network method is used to predict the ship traffic flow in Chengshantou traffic separation waters globally.Finally,based on the prediction of traffic flow in traffic separation waters,the navigation scene of commercial fishing ships in Shantou waters is constructed,and the ship navigation path optimization model is designed based on the improved genetic algorithm,which is applied to simulate the shortest safe path for commercial fishing ships when sailing in restricted waters.The research results show that within a certain error range,the traffic flow prediction model based on BI-GRU neural network has good robustness and generalization performance,and can realize global prediction of ship traffic flow situation in waters.The navigation path optimization model of commercial fishing boat based on improved genetic algorithm has good global optimization and convergence performance.The research in this paper has reference value for ensuring the navigation safety of the mixed navigation waters of merchant ships and fishing ships and reducing the collision risk of ships sailing in this area.