Study On Navigation Risk Identification And Accident Causation Analysis Of Autonomous Cargo Ships

With the rapid development and application of ship intelligent technologies,autonomous ships are believed to play a dominant role in the shipping industry in the future.In 2017,the International Maritime Organization?IMO?proposed the concept of Maritime Autonomous Surface Ships?MASS?at the 98^th Maritime Safety Committee?MSC?.After that,at the 99^th MSC,the preliminary definition of autonomous ships was officially given,along with the stages planned for the development of the autonomous ships.However,the research on autonomous cargo ships is just in its infancy,and many theories and models need to be tested and verified before their industrial application.In the early stage,the navigation of autonomous cargo ships suffers from different types of risks from various sources.For third level autonomous cargo ships,it is of great significance to carry out the navigational risk identification and causation analysis of the autonomous cargo ship accidents.In this thesis,different methods of text mining,fault hypothesis analysis,fault tree,and fuzzy theory were used.Modeling and analysis were conducted to identify the risk factors of transportation accidents of autonomous cargo ships,model the risks,and analyze related accidents for the human factors of Shore Control Centers?SCC?under remote control mode.The main research topics are summarized as follows.Identification of navigational risk factors of autonomous cargo ships.Based on the statistical analysis of 382 waterway transportation accidents of traditional ships,text mining was used to extract the risk factors.These factors were then analyzed and discussed through fault hypothesis analysis in order to identify the risks that may also occur on autonomous cargo ships.Based on the expert experience and the relevant literatures,some emerging risk factors faced by autonomous cargo ships were incorporated and analyzed.Modeling of navigational safety risks of autonomous cargo ships under autonomous navigation mode.Based on the identification of navigational risk factors of autonomous cargo ships,the fault tree model of autonomous cargo ships in autonomous navigation mode was constructed.By combining with the fuzzy theory,the key importance of basic events was calculated,and the minimum cut set was got.According to the minimum cut set,the contingency plan was selected,and the risk prevention and control measures were formulated according to the key importance of the basic events.It provides a theoretical guidance for the risk prevention and control of autonomous cargo ships under the autonomous navigation mode.The research results will also support the design of alarm systems in the future when the working mode needs to be switched to remote navigation mode under emergency situations.Modeling of human failures in SCC under remote driving mode.Through the analysis of the emergency treatment process of the SCC personnel,the human error event was divided into three stages of perception-decision-action based on the human error probability prediction method,and the three stages were modeled by constructing three Bayesian network models respectively.Base on the fuzzy theory,the expert judgements were utilized to obtain the basic probability of the root node and determine the conditional probability of each node in the Bayesian network,so that the total probability of human errors can be calculated.Finally,the importance of human failure factors were quantified by using a sensitivity analysis and some countermeasures were proposed accordingly.The recommendations will provide useful references for the construction of shore-side centers and the training of staffs.