Research On Intelligent Collision Avoidance Method Of Coastal Ships

Due to the increasing density of ships sailing in coastal areas,insufficient collision risk judgment from crew and unreasonable collision avoidance actions,etc.,These reasons cause ship collision accidents occur from time to time.At present,most of the research work related to ship intelligent collision avoidance studies how to avoid dangerous ships,ignoring the risk caused by the untimely update of ship navigation data,and the implementation of collision avoidance strategies takes less consideration of navigation rules and actual ship handling habits of ship crew.This paper is aimed at the coastal open sea,conducts in-depth research on related technologies including ship track prediction,collision risk evaluation,encounter situation judgement and collision avoidance algorithms in the field of navigation,and proposes a ship collision avoidance decision-making framework based on improved cultural particle swarms.This framework gives the whole process of the coastal ship’s intelligent collision avoidance method,and clarifies the functions that the ship’s intelligent collision avoidance method should have.First of all,the short-term prediction of the ship’s trajectory is combined with the Kalman filter algorithm.Based on the short-term prediction data,the error caused by the untimely update of the navigation data can be avoided.The ship collision risk evaluation model is used to calculate the predicted trajectory data to solve the quantified collision risk value.The encounter situation corresponding to the most dangerous ship is analyzed,and refer to the relevant collision avoidance rules under the specific situation to clarify the direction of the ship’s rudder angle.Then,the adaptive and heuristic methods are used to improve the cultural particle swarm algorithm,the adaptive ability of particles is enhanced from two aspects,including particle evolution speed and particle aggregation,and the collision risk model and ship rudder angle direction decision model are added into the fitness function of the particle swarm,the problem that the collision avoidance strategy is inconsistent with the actual navigation is solved.By comparing with the search process of other collision avoidance algorithms,the improved algorithm has better global convergence,and the convergence speed and stability are also significantly improved.Finally,the method is applied to the simulation environment of electronic chart.The part of experiment designed three basic encounter scenarios for single-ship encounters and three-ship encounters to verify the method proposed in the framework of this article.The results show that the method can realize ship collision avoidance and conform to the ship navigation rules and the pilot’s actual ship handling habits.To a certain extent,it improves the efficiency of ship collision avoidance and the safety of ship navigation.