Contrasting Track Out Of Control Ship Based On SVM Bridge Waters

With the rapid development of China's economic construction, a large number of high investment, new technologies, large-span of the Sea Cross River Bridge have already been built up. Only with Tongjiang-Sanya Sea online on the proposed five projects, namely Bohai Sea project, Yangtze River Estuary Sea project, Hangzhou Bay Bridge project, the Pearl River Estuary Lingdingyang works projects and Qiongzhou Strait. In addition,Zhoushan Islands Link Project, Qingdao to Huangdao, and the Yangtze River, Pearl River, Yellow River etc, a large number of bridge engineering. According to statistics, until March 2005, only the section of the Yangtze River in Yibin to Shanghai have been opened to traffic in the large bridge 41, in the construction of the bridge 18 (not including the planning or design of the bridge). So many bridges, not only bring in huge national and local economic profits, but also affect shipping safety.Nowadays,Shipping with the development of the navigation of ships became bigger, faster, and more, while the bridge area changes in the environment, such as velocity, wind speed, navigation, erosion, siltation, tidal and so on, will affect the safe navigation of ships. Dangerous situation in which the ship out of control in water traffic safety has always been a major hidden factor, due to ship out of control and triggered a major maritime accidents, the total number of accidents occupy a large proportion, and once the ship to lose control of the bridge waters occurs the consequences to a ship hit the bridge than the normal of the collision is much more serious. So out of control for the bridge area to predict the ship's track became more and more engineering concern.Firstly, sum up the track background and prediction of ships which out of control in the area of bridge waters, purpose and significance is illustrated on the research status were reviewed, and research papers ideas and methods elaborated explanation.Secondly, the use of systems engineering analysis of the impact of the relevant principles of the ship trajectory control many factors that control the ship track for the forecast model provides a predictable basis.Finaly, the paper support vector machine (SVM) the development, concepts and theories have been described, and then leads to SVM network, SVM algorithm described the mathematical description of the network used to predict based on the principle of SVM is proposed prediction step and forecasting feasibility. And then discussed in detail based on the SVM network in the area nearby out of control ship trajectory prediction model. Include the following aspects:the establishment of SVM-based forecasting model of key network technologies, including sample selection and pretreatment, the selection of input and output variables, training algorithm and parameter selection, the last a reasonable network model.