一种利用海量AIS数据规划安全航线的方法

为了解决智能规划的船舶航线贴着障碍物、岸边等问题，提出了一种基于海量AIS数据，研究船舶习惯航路并以此规划航线的方法。将分析水域网格化，选取目标船舶，统计其航迹所占的网格并计数，形成单船的航迹网格图。叠加不同目标船的航迹网格图，可获取分析水域内的航迹分布网格频数图，该图凝结了前船行驶的习惯航线。依据前船习惯航线设置网格的适航度，利用适航度改进A星算法的估值函数，可规划两点间安全航线。利用成山头警戒区作为分析水域，构建该水域3个月内100~150 m船舶形成的航迹分布网格频数图，并以此为依据规划航线，结果显示推荐的航线倾向于选择适航度较高的网格，有效避免以往研究中航线贴着障碍物、岸边的情况。

Study on method of safe routes planning with massive AIS data

In order to solve the problem that the intelligent planning ship routes are close to the obstacles and shores, a method of studying the customary routes of ships and planning the routes on this account based on massive AIS data was proposed. In this method, the gridding of analysis waters is carried out, and after selecting a target ship, the grids occupied by its track are counted, then the track grid diagram of single ship is formed. Through superposing the track grid diagrams of different target ships, the grid frequency diagram of track distribution in the analysis waters can be obtained, which blends together the customary routes of former ships. The navigability degree of each grid can be set according to the customary routes of former ships, and the evaluation function of A-star algorithm can be improved by using the navigability degree, then the safe routes between two points can be planned. Taking the Chengshantou warning zone as analysis waters, the grid frequency diagram of track distribution formed by 100-150 m ships in the analysis waters during three months was constructed, and the routes were planned on this basis. The results showed that the recommended routes tended to select the grids with higher navigability degree, which can effectively avoid the situation that the routes have been close to the obstacles and shores in previous research.