Analysis Of The Characteristics Of Ship Collision Avoidance Behavior In Sight Of One Another

Ship collision accidents have always been one of the most common types of marine traffic accidents and often have serious consequences,property losses,and marine pollution.According to the survey statistics,ship collision avoidance behavior largely determines the probability of ship collision.In this context,to study characteristics of collision avoidance behavior under ship encounters is of both theoretical and practical significance of ship collision avoidance decision-making,ship traffic simulation research,ship traffic management,and navigation risk assessment.AIS data contains massive,real-time ship dynamic and static information,which can effectively solve the problems of difficulty in collecting ship data,a small amount of data,and subjective analysis results.In recent years,it has been widely used in the research of ship behavior.Based on actual AIS data,this dissertation uses statistical analysis methods and the support vector machine regression models to analyze the characteristics of ship collision avoidance behavior under different encounter situations.Specific research work includes:(1)Sample extraction of ship encounters based on AIS data: After analyzing and preprocessing the AIS data,effective static and dynamic information of ships and relative motion parameters between ships are obtained.Based on the analysis of the division method of ship encounter situation and the actual situation of ship collision avoidance,the method of extracting the sample of ship encounter from AIS data is put forward.Taking the AIS data of Qiongzhou Strait as the research object to conduct empirical analysis,and samples of ship encounters under different encounter situations are extracted to provide data support for subsequent research on ship collision avoidance behavior.(2)Statistical analysis of ship collision avoidance behavior: After further processing of the ship encounter samples,the ship collision avoidance behavior data are extracted.The opportunity and range of collision avoidance are analyzed in detail by using the statistical analysis method,and the results are compared with the results of questionnaires.The consequence indicated that the actual opportunity of collision avoidance is late and the range of collision avoidance is small.(3)Estimation of ship collision avoidance behavior and weight analysis of decision factors: The ship motion parameters closely related to collision avoidance decision are selected as input feature vectors.The degree of course alteration and the range of speed variation are selected as output vectors.The mapping relationship between ship motion parameters and collision avoidance behavior obtained by the model proves the feasibility of using ship motion parameters to estimate collision avoidance behavior.By analyzing the weights of various factors in different encounter situations on collision avoidance behavior,it is concluded that the relative distance between the two ships and the ship speed ratio have the greatest influence on the collision avoidance behavior.This dissertation aims to quantitatively analyze the characteristics and laws of actual collision avoidance behavior under different encounter situations based on AIS data;it provides a reference for ship collision avoidance decision-making,ship traffic simulation,and ship traffic management.