Dynamic Programming을 이용한 무인잠수정의 최적경로생성에 관한 연구

Abstract

In contrast with man-operating underwater vehicles, autonomous underwater vehicles (AUVs) should find a proper trajectory, should control to follow a trajectory, should avoid static or moving obstacles, and should home safely to their mother ship or launching device by themselves. Due to these kinds of newly introduced problems, there are many challenging works to be solved compared to man-operating underwater vehicle cases. In these works, we focus on homing problems. Most homing devices are designed to have a shape of cone, so homing process could succeed when AUV has specific body attitude in front of the homing devices. And Most of AUVs have underactuated or nonholonomic characteristics. Those aspects have caused some difficulties in AUV's homing. One branch of studies of solving homing problem is path planning before taking actions. This paper addresses the optimal homing path planning for an AUV in planar space based on Dynamic Programming. And to redress shortcomings of Dynamic Programming which has been criticized as an impractical path planning method, we introduce two concepts, regulated nodes by maximum surge speed and solution index. To show the practical aspect of proposed path planing method, simulation results will be presented in this paper.