Based On The Fuzzy Pattern Pattern Recognition Of The Collision Risk Evaluation Method

With the dovelopment of shipping industry, Ship number has increased dramatically. Therefore adds to the ship collision accident, Which also bring our country casualties and huge economic losses. To ensure the safety of navigation and better serve to our national economic development, reducing the accident of ship collisions is important.Currently, at the practice of ship collision avoidance, it mainly relay on the seamen's subjective decisions. The accuracy about completing collision avoidance task with rules of thumb is to be strengthened. How to effectively avoid the collision, how to complete the Automation and intelligent of the collision avoidance systems. witch in order to reduce the labor intensity and seamen's psychological burden.so far. There is no effective way to make staff feel generally satisfied. Meanwhile, it is a hot topic in navigation session at home and abroad nowadays.In recent years, the period of study and writing papers, the author has been tracking, analysis and study the latest achievements about the ship's collision risk index, and do research for the establishment the foundation of supporting collision avoidance systemThis paper focuses on discussing the building of Ship Collision Risk Index mathematical model, the focal point is to define the influence of ship collision risk index, to get the membership function, and to judge the index level based on fuzzy pattern recognition theory. Through the study research and summarized of the collision risk index and fuzzy pattern recognition theory, combined with expert's opinions and my own maritime practice. This paper summarizes the analysis on the basis of reference to determine the impact of the collision risk index of the four main factors, and define three levels based on it. On this foundation, the model is from fuzzy pattern recognition theory, and defines the sample index level by this way.By analysising the results of the validation sample. Proved it to be feasible to use fuzzy pattern recognition the risk collision index.