Research On Path Planning Algorithm Of Unmanned Surface Vehicle Under Dynamic Water Depth Environment

With the development of unmanned technology,unmanned surface vehicle were increasingly used in water quality monitoring,submarine survey,marine cruising and marine transportation.In order to ensure the safe navigation of unmanned surface vehicle in the process of completing tasks,it was of great importance to carry out path planning research.In order to conduct the research on the path planning of unmanned surface vehicle in dynamic marine environment more accurately,this paper constructs a dynamic environment model based on the water depth data and tidal data of electronic charts.In the planning and analysis phase,two steps are divided into global dynamic path planning and local path planning.In the global dynamic planning,the improved LT-D~* Lite algorithm is used for solving the large-scale raster paths.For local path planning,a hybrid planning algorithm combining the LT-D~* Lite algorithm and the dynamic window method is used for local path planning.The main research contents are as follows.First,the path planning environment model is constructed.For the problems faced by the unmanned surface vehicle in global path planning,such as the lack of environmental information and the difficulty of establishing the environment model in real time,the static raster bathymetric environment is generated based on the discrete bathymetric data extracted from the vector electronic navigation charts and calculated by interpolation of the spline interpolation containing obstacles.On the basis of the static bathymetry model,the tidal data is superimposed to project the real-time changing water level to form a dynamic real-time bathymetry model.Secondly,the traditional dynamic algorithm has many turning points,difficult to track and control,and easy to ignore the influence of dynamic bathymetry environment and other shortcomings.According to the dynamic ocean environment model,the heuristic cost function is improved by establishing the bathymetric hazard cost function to select the optimal safety node and balance the safety of the path;on the other hand,the line-of-sight checking algorithm is introduced in the process of expanding the node to expand the node from eight neighborhoods to any angle and make the path smooth.The path is adjusted by combining the new obstacle environment after the tidal change.The simulation experiment verifies that the improved LT-D~* Lite algorithm has the characteristics of short and smooth path planning,and it can significantly reduce the risk of water depth and guide the unmanned surface vehicle to avoid the shallow water depth area.Finally,in the local planning,in order to improve the shortcomings of the dynamic window algorithm that the path is long and deviates from the target point when it is close to the target point,the global path is segmented according to the key nodes of the LT-D~* Lite algorithm to establish the local target point,and then the local path is planned by combining the dynamic window algorithm.The comparative simulation experiments verify that the hybrid planning algorithm improves the success rate of reaching the target point and achieves the optimization of path length,smoothness and security.The path planning algorithm research done in this paper has a catalytic effect on the autonomous navigation of intelligent ships and provides a good foundation for China to conduct several maritime missions with an intelligent platform.