Research On The Navigation Risk Model In Malacca And Singapore Straits Based On Stepwise-regression

Malacca and Singapore Straits is the traffic fortress of Indian Ocean and the Pacific, it is also the throat connecting Asia, Africa and Europe, the transport of goods accounts for a third of the world’s total trade, about80%of the imported crude oil transportation is realized through the straits in our country. With further prosperity of the Asian economy, ships, which cross the straits or dock in Singapore port, will significantly increase and this will inevitably affect the navigation safety and environmental safety in this water area. A scientific research on risk analysis for the maritime traffic safety of this straits will help to prevent and reduce safety accidents in this water, safeguard the maritime "lifeline".The safety of navigation is affected by many factors, it’s difficult to gather various aspects of information and complex to synthesize models including all the factors. Therefore, in order to reduce the number of independent variables, this paper uses the stepwise regression method to obtain the key factors affecting navigation safety, and establishing a regression mode for the subjective navigation danger degree.Firstly, this paper review the domestic and foreign research relate to navigation risk analysis, put forward the research content and method, then introduces the basic theory of regression analysis and questionnaire survey. Secondly, according to the report line, the straits is divided into nine sectors, we collected the data of subjective navigation danger degree of each sector and17factors which influence the safety of navigation in this Straits by questionnaire, analyze the factors one by one; then calculate the subjective navigation danger degree of each sector. Finally, based on the subjective navigation danger degree as the dependent variable, the frequency of occurrence for17affecting factors leading the risk of navigation as independent variable, we use stepwise regression method to establish the regression model of subjective navigation danger degree. At the same time, this paper carried on a quantitative research for four independent variables:visibility, background light, traffic congestion degree, fairway and anchorage planning status. By compared to other research, the applicability of the model was verified.