Ship Collision Risk Analysis And Research

Ship collision accident rate is always very large, ship collision usually brings vessels and their equipment damage, goods loss, casualties, pollution, and so on. The accident brings irreparable damage to human life, property and environment. So analysis ship collision risk in waters, understand ship collision probability and the possible consequence, make the corresponding measures, minimize the risk as far as possible is very important.The theses first systematically summarizes ship collision risk of home and abroad, basically grasp the current situation of ship collision risk analysis and the areas in need of improvement, in ship collision probability calculation, this thesis adopts the artificial neural network method.By collecting data and having an interview with relevant experts, grasp 30 domestic coastal ports hydrology, meteorology, waterway and traffic information, and so on, select visibility, wind, velocity, channel width, weighted traffic, navigation marks six factors as artificial neural network input data, before training,30 samples data are standardized and normalized in order to eliminate the influence of level and dimension, through the training of 800 times, obtain the ship collision probability of the application of artificial neural network, and check with the actual probability sample, and the result is satisfactory.The thesis establishes the framework of ship collision risk analysis with the basic of probability risk assessment method, sets up the mathematical model of ship collision risk analysis, the 30 coastal ports ship collision risk is analyzed. When evaluating the ship collision consequence, quote the concept of consequence equivalent, in a certain extent, solve the problem which ship collision indirect consequence is difficult to quantify, but because of ship collision consequence may be affected by ship, goods, person, environment and society and so on, the quantification of consequence is one of the obstacles in ship collision risk, therefore, thesis has a lot of further research value.