An UAV-assisted Dynamic Vessel Guidance Route Planning Method for Intelligent Port Navigation Considering Real-time Environmental Evolution

Abstract

Port traffic changes on a time scale that is poorly served by fixed sensors and precomputed routes. We therefore formulate vessel guidance as a repeatedly updated planning problem and use an unmanned aerial vehicle (UAV) to supply local observations when conventional sources are delayed or incomplete. Electronic navigational chart (ENC) constraints, automatic identification system (AIS) reports, and UAV detections are registered in a time-indexed representation of the navigation area. This representation drives an ECNA-based collision-risk term within the route optimizer, while a feedback loop revises only those route segments affected by new observations. In the simulation case, the resulting route is 0.57% longer than the geometric shortest path but is smoother and remains computable at millisecond scale. Adding UAV observations raises detection coverage from 72.0% to 96.4% and lowers the reported collision-risk index from 0.38 to 0.12. These results indicate that mobile aerial sensing can make port guidance more responsive without sacrificing online computational feasibility.