다물체 해양로봇의 실시간 해석을 위한 모델링 및 시뮬레이션 기술 개발

Abstract

Remotely operated vehicle (ROV) type underwater robots are used for precision work and exploration on the seabed. The operator of this type of robot has to depend on an optical camera to work, which causes problems such as low workability and high operator fatigue. In particular, for heavy work underwater, tracked vehicle robots disturb floats on the ground due to contact with the seabed when driving, making it difficult to work. To solve these problems, a Cyber Physical Operation System (CPOS) is being developed. This paper deals with modeling and simulation technology that can analyze the real-time dynamics of underwater robot, which is one of the core technologies of CPOS. As offshore operations, force feedback to the operator can increase the stability and efficiency of the robot and the operation. In addition, the mechanical stability of the system can be monitored by estimating the effect of the reaction forces on the joints of the system. For this purpose, we will estimate the reaction force using the deep learning algorithm by replacing the force sensor, and estimate the joint reaction force of the whole machine model using multibody dynamics analysis technology.