Maritime Traffic Accident Research In Dalian Jurisdiction

On the basic of historical data of maritime accidents in Dalian jurisdiction from2001to2012, this dissertation makes reasonable analysis and research by applying the maritime statistical analytic method and self-adaptive filter prediction model respectively. It gives a scientific definition to maritime accident and thoroughly studies the research status of this field. Consequently, it chooses the maritime statistical analytic method and self-adaptive filter prediction model in accordance with the characteristics of maritime accident. After classifying the type and level of maritime accidents based on the Statistical Methods for maritime Accident, the dissertation summarizes the characteristics of maritime accidents and analyzes the fundamental factors for causing maritime accidents from the aspects of shipping, environment, crew and management. Maritime accident is characterized by serious pernicious, huge losses, regularity, arbitrariness, contingency and small probability. By doing corresponding research on the characteristics of maritime accident, the unique regularity of Dalian maritime accidents can be found as following. The principal accidents of Dalian maritime accidents are minor accidents, which takes the proportion of70%and they mostly occur in the second quarter of every year. In all kinds of maritime accidents, the collision accidents, which mainly occur on general cargo ships, rank first with the proportion of50%. Meanwhile, the accidents happening in good visibility account for74%, far more than that in restricted visibility. In the aspect of region, the accidents concentrate in some special water areas such as the waters of the harbour areas, Dasanshan and Xiaosanshan waterway and Laotieshan waterway. And then the paper gives scientific and rational prediction about the maritime accidents in Dalian water areas. With the help of historical data from2001to2012about the maritime accidents in Dalian jurisdiction and the theory of self-adaptive filter prediction mode which can be automatized through the C++language program, the dissertation firstly makes prediction and error analysis on the different kinds of maritime accidents in Dalian jurisdiction and gets the conclusion. The conclusion says that whether the yearly relative error of all kinds of accidents or the totally relative error of accidents happening in these two years is less than0.5. Moreover, the prediction error for common and serious accidents in2011and that for the serious accidents from2011to2012reach to0. For this reason, the prediction precision of this model basically meets the need of efficacious prediction for Dalian maritime accident.Finally, the scientific prediction for the2013maritime accident in Dalian jurisdiction is made and the relevant suggestions are offered in this dissertation.