| The fourth industrial revolution represented by cloud computing,Internet of things,artificial intelligence,machine learning and other application technologies is affecting our world.The development of communication technology and the improvement of information analysis capabilities have provided revolutionary development opportunities for various industries.In the shipping industry,autonomous ships are becoming more and more concerned because they can transport goods and passengers in a safer,more efficient and environmentally friendly manner,which has become one of the future development directions of the shipping industry.The operation of autonomous ships is based on many emerging technologies,and autonomous ships are free from human intervention during operations,which will inevitably lead to new risk factors that will affect the success of autonomous ship missions.Some foreign countries and organizations have carried out risk assessment research on autonomous ships,but more general risk assessments have been conducted on the entire voyage cycle of autonomous ships from a macro perspective,and there is a lack of risk assessment for specific autonomous ship operating tasks.In this thesis,the risk analysis and early warning method research is carried out in the berthing and unberthing operation stage,which is difficult to maneuver in the entire navigation cycle of autonomous ships,in order to improve the active safety during berthing and unberthing operations of autonomous ships and avoid the occurrence of operational accidents.The main research results are as follows:(1)The task flow of autonomous ship berthing and unberthing operation is analyzed.Based on the comprehensive consideration of the responsibility of system equipment involved in autonomous ship berthing and unberthing operation,the risk factors of autonomous ship berthing and unberthing operation are identified by task FMEA method.As a whole,19 observation risk factors including natural environment risk factors,ship state risk factors and hardware equipment risk factors are identified,and the influence of risk factors on autonomous ship berthing and unberthing operation is analyzed in detail.(2)Based on the identification of risk factors and the analysis of the influence of risk factors,the risk evolution analysis of typical operation risk accidents of autonomous ship berthing and unberthing operation is carried out,and the risk evolution model structure is constructed by static Bayesian network.On the basis of risk evolution analysis,the time correlation problem of risk scenario evolution with time process is dealt with.Considering the time dependence relationship between risk factors,the static Bayesian network topology model of risk evolution is expanded into corresponding dynamic logic gates of different time slices according to the time axis.The dynamic Bayesian network structure of autonomous ship berthing and unberthing operation is established,and the probability distribution of root nodes and leaf nodes of network model is carried out to complete the establishment of risk analysis model of autonomous ship berthing and unberthing operation.The scene data of autonomous ship berthing and unberthing operation are set to simulate the process of autonomous ship berthing and unberthing operation,and the risk analysis model is driven to calculate the risk.(3)Based on the risk analysis of autonomous ship berthing and unberthing operation based on dynamic Bayesian network,the dynamic risk early warning method of autonomous ship berthing and unberthing operation is proposed.The risk early warning process and data flow of autonomous ship berthing and unberthing operation are introduced.The risk early warning software of autonomous ship berthing and unberthing operation is developed by MATLAB The purpose is to embed the autonomous ship operation control system,which can improve the active safety and avoid the occurrence of operation accidents in the process of autonomous ship operation.Simulate the scene of autonomous ship berthing and unberthing operation environment,and introduce the operation process of the developed risk early warning software through examples. |