조류 영향을 고려한 강화학습 기반 해양운동체 경로 최적화

Abstract

This paper proposes a global path planning algorithm for a marine vehicle considering the vehicle’s dynamic characteristics and the disturbance effects in ocean environments. Unlike path planning on land, various environmental disturbances such as winds, waves, currents should be considered in path planning in ocean environments. Among the disturbances, the effects due to ocean currents are primarily considered in this study. A dynamic model is used to simulate realistic motion of the vehicle, which does not allow the use of conventional search-based optimization algorithms such as A*. Instead, a reinforcement learning algorithm is employed for path optimization in the study. The proposed algorithm provides a near optimal path between two points when the map and the current field data are given.