Research On Route Planning Technology Of Inland Water Network Oriented To Intelligent Navigation

There are many inland river systems in China,and navigable waterways are criss-crossed,and inland river shipping is very developed.However,due to the complex and changeable inland waterways,some areas even formed dense water networks,which caused great trouble to the navigation safety of navigable ships,and was prone to accidents such as stranding and bridge collision.In recent years,the construction of inland digital waterways has greatly improved the cognition ability of ships on the waterway environment from the data level through channel perception and digital methods.The gradual popularization of electronic chart systems has also provided powerful power for ships from the application level.Auxiliary navigation means.However,the current electronic chart system has played a more role in navigation monitoring,and still does not have the ability of automatic route planning,it is difficult to help ships plan ahead to avoid risks and improve navigation efficiency.In response to this problem,this paper studies the automatic route planning technology of the inland water network based on the channel data basis of inland river chamel perception and collection,combined with the ship's own conditions,and using artificial intelligence methods,and provides services for the ship through the navigation aid application based on the electronic channel mapto assist ships to avoid accidents such as entering a wrong route and hitting bridges,and improve the efficiency of navigation by providing optimal routes to lay the foundation for intelligent navigation of ships.The main research work completed in this paper includes:(1)According to the characteristics of the cross-channels,the directed topology network data structure of the inland water network is designed,and the A star algorithm is used to improve from three aspects:adding channel constraints,optimizing the cost function and valuation function,and a route planning algorithm adapted to the navigation conditions and morphological characteristics of the Inland River channel is proposed.(2)Aiming at the inland waterway traffic separation system,a hierarchical route planning method based on A-star and DBSCAN density clustering is proposed.The route planning process is divided into two parts:high-level route planning and low-level route planning to improve the accuracy and effectiveness of route planning.(3)Using the idea of statistical processing of ship traffic flow by segment,a spatio-temporal neural network model combining the dynamic and static characteristics of the channel is proposed to achieve accurate prediction of the sailing time of the ship on the planned route.(4)Design the navigation system of inland ships,use TensorFlow-serving to realize the service interface of route planning and sailing time prediction,and implement the application of route planning in the APP,WeChat applet and other clients.Based on the characteristics of the inland water network,this paper focuses on the intelligent navigation of ships and deeply studies the methods of automatic route planning and sailing time prediction.It is a useful exploration of artificial intelligence in the field of ship navigation.It not only has theoretical significance but also has good application value.