Research On Human Factors Reliability Of MASS Based On Improved BN-CREAM

The research of Maritime Surface Autonomous Ships(MASS)has become a hotspot in recent years.The development of automation technology has broadened the application prospects of MASS,and inspired a series of research and discussion related to the development and operation of MASS.MASS based on the new ship operation concept is expected to improve operation efficiency and higher reliability,reduce operating expenditure and greenhouse gas emissions,and bring safety and economic benefits to the entire maritime industry.In order to ensure the safety operation of MASS,the Shore Control Center(SCC)will monitor and intervene the MASS.Although MASS has less involvement in human factors during navigation,it will still rely on operators working in the onshore control center under certain circumstances.Therefore,this paper will use Human Reliability Analysis(HRA)methods as the basis to study the task events and failure probability of SCC operators in MASS collision avoidance events.The results of this paper will provide valuable information for the design of regulations and plans of autonomous ship collision avoidance systems.First of all,according to the theoretical framework of Cognitive Reliability and Error Analysis Method(CREAM),the process of people’s receiving external information is divided into four stages: observation,interpretation,planning and implementation.The MASS human error model for collision avoidance is established based on team information,decision,and action in Crew context model(IDAC)and CREAM retrospective analysis method,Analyze the root cause of human error in MASS collision avoidance scenario.Secondly,the CREAM basic method was improved.The Grey-DEMATEL method was used to explore the correlation between CPC factors,and the factor weights were obtained.The study found that “working conditions”,“number of simultaneous targets”,“available time”,and “quality of team members cooperation” were more affected than other factors,“Completeness of organization” is the most critical CPC factor that causes human error events in MASS collision avoidance operations,while the weight of “duty interval” is the smallest,which has less obvious impact on the operation.Then the factor weight and expert evaluation results are introduced into Bayesian network reasoning,and the results show that the operators are mainly in the “strategic” and “tactical” control modes when conducting MASS collision avoidance operations again.Then,combined with Hierarchical Task Analysis(HTA),the whole process of MASS collision avoidance operation was explored,and the hierarchical task analysis diagram of MASS collision avoidance operation was obtained.Combined with the task sequence,the CREAM expansion method was used to analyze the collision avoidance process,and the weight factor evaluation of the four cognitive functions in the CREAM expansion method was improved.Finally,the probability value of quantitative analysis of human reliability of each task sequence was obtained.Finally,a weighted BN-CREAM method for human error analysis of MASS operators was established to conduct quantitative prediction and qualitative analysis on the probability of human error of SCC operators and the retrospective analysis,so as to improve the reliability of MASS operations.