Research On Data-Driven Route Planning Algorithm For Inland Waters

The inland waterway network represented by the Yangtze River system offers a natural advantage for the development of inland waterway transport in China.With the rapid development of the inland waterway industry,the need for intelligent inland waterway construction is becoming increasingly urgent.It is vital for transport ships to identify routes with good safety and economy before departure.In practice,however,there are still many vessels that determine their routes based on the captain’s own experience.As the navigable section of a river can change with the onset of flood and drought periods.If the captain does not have access to accurate information,there is a high risk to navigation.Therefore,based on the above,the study in this paper extracts the actual navigational position of a vessel from AIS data and uses a route finding algorithm and an optimization algorithm to improve the economics of the shortest navigational distance route as an optimization objective to solve the inland waterway route planning problem in local waters.The research content and results of the work mainly include.(1)Creating a mathematical model for inland waterway route planning.A raster map model based on the Mercator coordinate system is created,a river is divided into navigable and non-navigable zones using electronic river maps,and a discrete route model is created.A model of data-driven inland navigation planning algorithm based on the four stages of "hold-group-search-route-optimization" is created.(2)A multi-objective method of maximum DP compression for curved inland waterways is proposed.Based on the classical DP compression algorithm,this paper proposes a multipurpose maximum DP algorithm(MPDP)which includes four improved strategies and is more suitable for curved inland waterways,considering the drawbacks of DP algorithm in terms of poor compression of multi-turn trajectories,not considering two important factors of vessel manoeuvring such as course and speed,and not considering the correction of compressed routes together with maps.Simulation results show that the MPDP algorithm is optimised for several performance measures such as length loss rate,simultaneous Euclidean distance,average speed deviation and average course deviation.(3)A* quadratic algorithm based on clustering results,suitable for bifurcated inland waterways,is proposed for route search.In this paper,the DBSCAN algorithm is used to cluster the compressed key points of a waterway.A method is developed to determine the relationship between waypoint clusters based on clustering results,and a directed graph of waypoint clusters is constructed.A* quadratic algorithm is developed to generate recommended routes and alternative routes for the captain to choose according to the situation of multiple island bifurcations on inland waterways.The solution route of the quadratic algorithm A* is verified using simulation experiments and the shortest route is successfully generated as the recommended route and the alternative route,which is far from the shortest route and better.(4)A route optimization method based on sine-cosine particle swarm algorithm is proposed.In this paper,the particle swarm optimization algorithm is used to optimize the global route using the shortest navigational distance as the target.Based on the particle swarm algorithm,a sine-sine particle swarm optimization algorithm(SCPSO)with extended random particle perturbation is proposed to improve the search capability of the particle swarm algorithm,and the effectiveness of the algorithm is tested through simulation experiments.(5)Building a simulation system for inland waterway planning.In this paper,a simulation system for path planning has been created using MATLAB App Designer.Based on this simulation system,the compression algorithm(MPDP algorithm),clustering algorithm(DBSCAN algorithm),route search algorithm(A* quadratic algorithm)and optimization algorithm(SCPSO algorithm)used in the algorithm are simulated and experimentally tested in the context of two water areas of Yangtze River inland waterway network.Experimental results show that the algorithms proposed in this paper have good route planning capability.