The Ship Capsizing Risk Assessment Based On BP Neural Network

Ship as an important carrier of maritime transport is a complex and bulky system engineering,at the same time its "high risk" has attracted more and more scholars to study its risk control.And with the development of big data and artificial intelligence,it breaks down the predicament of complex risk assessment due to difficult data.In this paper,the basic theory of artificial intelligence-neural network,is used for the assessment of the risk of ship capsizing,which has serious consequences of the accident.This article analyzes the risk factors of navigation and combines the risk connotation to calculate the risk from three aspects:the risk of hazard,the vulnerability of the disaster-affected and the recovery.Considering the marine meteorology,the status of the ship,the condition of the cargo,the crew and external rescue and other aspects,build a ship-capsizing risk assessment system.The system includes 23 evaluation indexes.In this paper,each index has been analyzed in depth,and each index has its own quantitative or qualitative methods.In the example,"Xin Bao Hang 87" is taken as an example to realize the quantification of indicators.Finally,according to the principle of BP neural network,the network structure of ship-capsizing risk is constructed,and 23 risk evaluation indexes are taken as input values,50 groups of sample data are queried and compiled,and two schemes are designed to train and fit BP neural network using Matlab.It is found that the scheme that already has a certain network structure fits well.In addition,BP neural network output value compared with the actual value,the error in the acceptable range,proved the effectiveness of the network.This fitting result can be used to predict the value of capsizing risk of other ships.This paper studies the risk assessment of ship capsizing based on BP neural network.It shows that there are many indicators that affect the risk of ship capsizing and the internal relations are complex.BP neural network eliminates the complex network by learning data,a major difficulty in risk assessment,practical.