Research On Emergency Response Decision-Making For Emergency Events In Sea Lanes Based On “Scenario-Response” Mode

Maritime transportation is a complex system consisting of many elements,leading to the occurrence of unexpected events in the sea lanes is common due to the intrusion of various risk factors.Although the international maritime department through the introduction of some policies,laws and regulations,the use of some new technologies and other measures to reduce the occurrence of emergency events in sea lanes,however,the risk of some typical emergencies such as maritime accidents still exists,and often cause serious consequences such as property losses,casualties,environmental damage and so on.Therefore,an effective emergency response is particularly important in order to reduce the loss caused by emergency events in sea lanes.As the occurrence of emergency events in sea lanes is often far away from land,coupled with the complex environment of sea lanes,the whole emergency response process is full of uncertainty,and the traditional single mode of emergency response based on the pre-defined plan cannot meet the actual emergency decision-making needs.The “scenario-response”decision-making is a real-time decision-making mode based on the scenario,which can make up for the shortcomings of the previous emergency response decision-making mode,and it is of great theoretical significance and application value to carry out research on the “scenarioresponse” emergency response decision-making for emergency events in sea lanes.In this thesis,we analyze the occurrence and development characteristics of maritime emergency events and the “scenario-response” emergency decision-making mechanism based on typical maritime emergency events in sea lanes,and then use advanced technologies such as sparse learning,expert systems and risk decision theory to study the key issues of “scenarioresponse”-based emergency response decision-making in sea lanes.The main research contents and results are as follows:(1)Research on the identification of scenario elements for emergency events in sea lanes.Scenario elements are the basis for scenario composition and evolutionary analysis.Considering the existence of complex relationships and redundancy among the antecedent scenario elements affecting the state of the objective scenario elements,we first construct a complex network based on association rules that can portray the relationships between the antecedent elements,and then assessed the importance of the antecedent scenario elements.Then,based on the importance assessment,we combined the logistic regression model that can analyze the relationships between the antecedent scenario elements and the objective scenario elements with the adaptive sparse group Lasso.The adaptive sparse group Lasso logistic regression model for the identification of scenario elements is constructed to realize the identification of key scenario elements when the objective scenario element is a categorical variable.Finally,the feasibility and effectiveness of the proposed model are verified by using maritime accidents as an example.(2)Research on the scenario evolutionary reasoning for emergency events in sea lanes.Scenario evolutionary reasoning analysis can grasp the development trend of maritime emergency events,which is the basis for effective emergency response.Considering the uncertainty of scenario information and scenario state in the scenario evolution process of the emergency events in sea lanes,a belief rule base that can be analyzed in this environment is used to construct the reasoning model,in which a belief rule base subsystem is developed by combining mutual information to generate a random subspace in order to avoid the combinatorial explosion caused by excessive input,and the structural form of the belief rule is extended to solve the multiple output problem.Then,all the belief rule base subsystems are ensembled to build a scenario evolutionary reasoning model for maritime emergency events in sea lanes,which enables efficient scenario evolutionary reasoning in an uncertain environment.Finally,based on the analysis of the cases and results of the scenario element identification,the feasibility and effectiveness of the proposed model are verified.(3)Research on the dynamic selection of emergency response alternatives for emergency events in sea lanes.Dynamic selection of emergency response alternatives is an important initiative to achieve effective scenario-based response.Considering the environmental complexity of the decision-making of the emergency response alternative for emergencies in sea lanes based on scenarios and the necessity of expert participation,the probabilistic linguistic term sets are first used to express the assessment of decision-makers in a complex decisionmaking environment,and a probabilistic linguistic term sets dissimilarity measure model that can retain the original framework of the probabilistic linguistic term is constructed based on the Jensen-Shannon divergence.In addition,for the bounded rational behavior of decision-makers in an uncertain environment assessment process,prospect theory is introduced and combined with the proposed probabilistic linguistic term sets dissimilarity measure model to construct a static rating computational model based on expert assessment information.Then,a dynamic rating calculation model that can consider the correlation between assessment information in different emergency decision stages is constructed to realize the optimal emergency response alternative selection and adjustment for emergency events in sea lanes.Finally,the feasibility and effectiveness of the proposed model are verified by the analysis of a case of maritime accidents.The “scenario-response” decision-making for emergency response in the sea lanes studied in this thesis involves the identification of scenario elements,scenario evolutionary reasoning,and dynamic selection of emergency response alternatives,which on the one hand enriches the application scope of the “scenario-response” mode and promotes its development,and on the other hand provides new ideas and solutions for emergency response decision-making of emergency events in sea lanes.