Study On Marine Traffic Risk Prediction Based On Catastrophe Theory

As a pillar industry for economic development, the shipping industry bears over90%of the global cargo transportation. Meanwhile, high-risk has been a significant feature of the shipping industry and as a consequence, damages and losses resulted in various aspects are far more than that incurred in any other forms of transportation. With the rapid development of China’s economy and the ever-deepening global economic integration, the issue of how to effectively predict and reduce maritime traffic risks has become a worldwide concern and a difficult point as well.Carrying out risk predictions for maritime traffic is of profound significance:it contributes not only to the enhancement of marine transportation safety and prevention of accident occurrence at sea, but also to the risk reduction and marine environment protection. This research meets the major demands of the national economy development, national security and scientific development concept as well, which is to bring huge social, economic and environmental benefits.The general rules and approaches to evolution, assessment, prediction and control of maritime traffic risks are revealed in this dissertation based on systematic principles and by application of catastrophe theory. To establish an ideological and theoretical framework of dynamics mechanism for maritime traffic safety system is also proposed in this dissertation, which presents a new idea for maritime traffic risks research with profound theoretical and practical significance.In light of the maritime traffic safety system comprising the three elements (namely "Man-Ship-Environment"), the approach of employing catastrophe theory as a tool to study the maritime traffic safety system is established, and a system cusp catastrophe potential function structural model with universal application based on the catastrophe structure of this system is presented. Furthermore, bifurcation set equations which contribute to structural instability are also derived to characterize the dynamic nature of the evolution of maritime traffic safety system. By analyzing the system state, both the evolution curve of system characteristics and evolution model of maritime traffic safety, along with the accident evolution model and the risk evolution model, are defined. The cell mapping theory is applied not only to describe the formation, mutual transformation and mechanism of these three evolution models, but also is to be used for further maritime traffic risk evolution model research.In allusion to the uncertainties of marine traffic risks, such as relativity, dynamics, ambiguity and mutability, limitations of traditional risk assessment methods are analyzed. By way of traditional risk model to derive the essence of fuzzy comprehensive assessment, fuzzy catastrophe principle is for the first time used in the field of maritime traffic risks. The comprehensive assessment model based on fuzzy catastrophe of marine traffic risk is established on the basis of maritime transportation risk assessment system. For this model, only relative ranking of assessment indicatives needs to be taken into consideration, therefore the subjectivity of assessment is effectively overcome and accuracy and efficiency of the risk assessment improved.On account of time series scattered points which are composed of maritime traffic safety system evolution values combined with the advantages of gray theory, gray time series model of the maritime traffic risk is established. With the derived sharp point catastrophes structure of this gray time series model, predictable conditions of maritime traffic risk are defined. And on that basis, taking the catastrophe point, catastrophe time, catastrophe direction, risk-causing capability and risk-causing efficiency as research objects, the maritime traffic risk prediction model is built based on gray catastrophe. After data consistency examination Expert Questionnaire is utilized to carry out safety assessment on different vessels in a certain port and to make risk predictions for the involved ships. By comparing the prediction result with the fuzzy comprehensive assessment method, the scientific nature and advancement of the model is verified.