Global Path Planning Of Intelligent Vessels Based On A* Algorithm

As the main development direction of the current shipping industry,intelligent ship is a technology that uses artificial intelligence instead of human beings to drive ships.In order to provide path planning decision information for intelligent ships,it has always been the focus of intelligent ship research to find a path planning algorithm that can provide safe and economical optimal routes for ships.According to oceangoing ships global path planning problem,this thesis presents a path planning algorithm based on the A * algorithm.Adaptive improvement is made to the cost function of A* algorithm.Quadtree adaptive resolution grid model was used to construct the environment,and the ship’s turning characteristics were included in the node evaluation.Finally,the output path was smoothed,and the specific contents were as follows:(1)Quadtree structure is used to construct environment for ship path planning.Based on the Quadtree image segmentation technology,the gray-scale electronic chart is segmented with adaptive resolution,and the grid objects with different gray-scale,position,size,edge and resolution are obtained.According to the tree structure of Quadtree and grid object position relationship,the node neighborhood structure of raster object is constructed.The cost model is built in Quadtree adaptive resolution grid object and embedded into the improved A* algorithm to support node search in ship path planning.(2)According to the traditional A* algorithm,the evaluation function is established for the factors affecting the navigation safety of ships,and the candidate nodes are evaluated comprehensively.Grid objects with different gray values have different type penalty values.According to the characteristics of ship turning limitation,the reachable region of ship is established by using geometric relations,and the unreachable points of candidate nodes are eliminated.The distance penalty function is added into the cost function to control the safe distance between the path node and the obstacle.According to the flow field data and the distance between path nodes and obstacles,the penalty function of flow course and velocity is calculated,so the course and velocity of ocean current is also taken as an evaluation item in node search.(3)The output path is smoothed by using cubic Bessel curve.On the basis of improving the output path of A* algorithm,control points are found for each path segment according to the length and angle relationship of each adjacent path segments.The endpoints and the control points of each path segment are used to calculate its corresponding cubic Bessel curve,and finally the smoothed path with continuous curvature is output.The research results show that the global path planning method of intelligent ship based on A* algorithm proposed in this thesis can effectively ensure the navigation safety of the ship,improve the search efficiency,ensure the output path smooth and its curvature continuous,and has certain reference value and practical significance.