Route planning and optimization for maritime collision avoidance

National and international sea routes are increasingly becoming busier with the development of global trade parallel to the demand for maritime transportation. These developments increase the risk of collision at sea. Collisions are frequently caused by human error such as maneuver timing mistakes, risk assessment failures and deficiencies in strategies for collision avoidance. These human related factors reveal the importance of the automation and decision support systems in providing safety of navigation. Thus, a decision support system has been developed in this study that can be reference to the ship operators in the implementation of the collision avoidance action in case of encountering a situation involving collision risk at sea. This study aims to create a system to assist a navigator for planning the collision avoidance route of a give way vessel in accordance with the international regulations for preventing collision at sea. The study also aims to minimize the distance of the altered course to avoid collision, usually subjectively conducted based on personal experience and competence, by managing the process through a decision support system. Qualitative and quantitative methods have been used in the study. In the qualitative research process, the variable constraints in the mathematical model of the solution algorithm and the inputs of the experimental tests have been determined based on the findings obtained from experts through interviews. In the quantitative research process of the study, the problem solution has been reached with the developed solution algorithm (ColAv_GA) which has been formed by the Genetic Algorithm and Fuzzy Logic. The developed algorithm has been tested in virtual environment with bridge simulator and in real environment with autonomous surface vehicle (ASV) with satisfactory results. The output of this research is expected to contribute the safety of navigation and decrease the navigation costs. The developed algorithm can be used as a collision avoidance sub-module for autonomous ships and unmanned surface vehicles.