The Research On Mathematic Models Of Decision-making In Ship Collision Avoidance

The Research on Mathematic Models of Decision-Making in Ship Collision AvoidanceAbstractThe systematic research has been made in this dissertation on the establishment of the mathematic model of decision-making in collision avoidance, especially in the influencing factors and the degree of collision risk.Four main problems have been identified in the field of intelligent decision-making system for ship collision avoidance, which is also referred to as collision avoidance automation system. The author considered that further research should be undertaken in the four areas of the problems identified. The first problem is on how to choose proper influencing factors to establish a mathematic model in accurate response to the degree of ship collision risk. The second is on the overall appraisal of collision risk in multi-ship encounter situation. The third is on the specific application of ship domain in the intelligent decision-making system for ship collision avoidance. The last is on acquiring knowledge and establishing knowledge-base in the intelligent decision-making system for ship collision avoidance.The factors have been analyzed in determining the degree of collision risk between two ships, the mathematic model method has been established to calculate the degree of ship collision risk, in terms of selecting influential factors and establishing the mathematic model of the degree of ship collision risk. For the first time, the assumption was given in the dissertation that the degree of collision risk should indicate the possibility of collision occurrence, the degree of urgency of collision risk, and the easiness and hardness in collision avoidance. Five factors have been defined in determination of the degree of collision risk based on the analysis of influential factors; the new mathematic model has been established in calculating the degree of collisionrisk by using the Fuzzy mathematics method.In overall appraisal of collision risk of multi-ship encounters, the idea and method have been brought up, based on the research outcomes of the mathematic model of collision risk between two ships, to handle the situation of the multi-ship collision avoidance. The study has been also undertaken on determining the vessel of most possible collision in the situation of multi-ship collision avoidance, using the inferential method.In consideration of the application of ship domain in the intelligent decision-making system, it has been found that, after comprehensive research, the concept of ship domain in collision avoidance should reflect the concept of a distance, not the distribution of points of DCPA around the ship itself. In the process of decision-making for collision avoidance, the area of ship domain should be used to judge the degree of collision avoidance, to determine the course to alter, and the right time to resume the course. All other ships then could be kept clear of own ship domain in the heads-on situation.The intelligent decision-making system for ship collision avoidance has been established to check and confirm the rationality of the mathematic model of decision-making for ship collision avoidance. In terms of acquiring knowledge and establishing knowledge-base in the intelligent decision-making system for collision avoidance, the author has categorized the knowledge of ship collision avoidance based on the study of artificial intelligence, on the Collision Avoidance Rules, on the characteristics of collision avoidance and the common practice of ship maneuvering. The author has further established the knowledge-base on the basis of the provision of The Rules of Collision Avoidance, the experience of experts and the good seamanship.