The Journal of Japan Institute of Navigation · 1986 · 16 citations · 0 references
EngineeringComputer ScienceMarine EngineeringDecision AnalysisKnowledge RepresentationShip ManeuveringCollision AvoidanceCollision Avoidance ControlCollision Avoidance ManeuveringExpert SystemsFuzzy LogicKnowledge-based Decision SupportDecision Support Systems For Collision AvoidanceFuzzy Expert SystemsFuzzy Logic Collision AvoidanceFuzzy SetsManeuvering Decision-makingMarine Collision PreventionShip Collision RiskVessel Collision Prevention
Fuzzy reasoning as decision analysis is very actively being investigated. In this paper, the method of the fuzzy reasoning is applied to the decision making for collision avoidance maneuver in congested waters. The proposed model consists of two sub-model. One is for evaluating the risk of collision on the values of TCPA and DCPA. Another is for deciding the collision avoidance action in order to satisfy the following three goals with the order of priority; 1) to avoid the most dangerous target 2) to obey REGULATIONS FOR PREVENTING COLLISION AT SEA, 1972 3) to minimizng the loss of sailing time. The knowledges of ship's operator as to the judgement about the situation were acquired by a series of experiments using a simple radar simulator. The results of experiments and the comments of ship's operator were fully used to renew the knowledge base of the model. Finally the maneuver by the proposed model were found to be reasonable under the simulated encounters. This means that the model will be useful as a fundamental approach in order to construct the so-called expert system for collision avoidance maneuver.