Optimal Route Search Based on Multi-objective Genetic Algorithm for Maritime Navigation Vessels

Abstract

Ocean research requires regular collection of ocean data, wherein an autonomous robotic ship is usually used. However, in contrast to collecting land-based data, collecting sea level data face the following problems. First, robot ships are affected by sea surface winds, waves, and tides, with constantly changing strength and direction. Second, hull collisions must be prevented when multiple ships are working simultaneously. Third, given the limitation of the electric power of the autonomous sailing ship, the electric power consumption of the robot ship must be considered when collecting over a wide sea. Fourth, fixed obstacles, such as an island on the sea surface, must be avoided. Given such issues, no effective navigation route search system is currently available. In this work, a navigation route system for complex situations on the sea surface was designed on the basis of the actual situation. Clustering method was used to classify collection points according to distance based on the number of robot ships, and a multi-objective genetic algorithm was used to determine the optimal path for each classification.