Research On The Design Method Of Ship’s Offshore Route Based On AIS Data

Automatic Identification System(AIS)equipment and systems have become widely used and popularized in the shipping industry.The vast amount of AIS navigation data broadcast by ships can not only be used for the supervision and management of maritime administration but also contains rich characteristics of navigation behavior that are worth further exploration and mining.Experts and scholars hope to make full use of AIS data to improve traffic information services in the shipping industry.Route design plays a pivotal role in the technology of ship intelligent navigation.If AIS data can be used to provide more reasonable and effective guidance,it can fully ensure navigation efficiency and safety.This thesis proposes a data-driven method to design recommended routes using data cleaning and data compression to obtain ships’ characteristic trajectories.Data-driven routes are generated using consistent point clustering and similarity measurement for trajectories,in response to the shortcomings of current route design methods and AIS data application methods.The main elements accomplished in this thesis are as follows: First of all,the cleaning of AIS data is divided into static cleaning and dynamic cleaning,with routes with missing information and erroneous sailing status cleaned using current data cleaning specifications in the industry,and missing information filled in reasonably;Then the Douglas-Peuker(D-P)algorithm is improved for the characteristics of offshore routes,and the key nodes of each route are extracted to realize the compression of ship tracks;The DBSCAN method is used to cluster the two endpoints of each track,and tracks with similar endpoints are grouped into the same cluster of tracks;After getting the same cluster of tracks,the track similarity measurement based on Hausdorff distance is used to find the baseline in each cluster of tracks.The recommended route corresponding to each cluster is generated based on the baseline using a sector scan method,which achieves the goal of generating recommended ship routes based on trajectory.This thesis proposes a data-driven method to design recommended routes based on the operation of cleaning and clustering the AIS historical data of the Gulf of Mexico and its surrounding waters for the whole year of 2020.The corresponding recommended routes are obtained using the proposed route generation method,and the generation effect can meet the design requirements of offshore routes.This provides a reasonable research direction for the route design field of the shipping industry.