자율운항시스템 개발을 위한 선박 운동제어에 관한 연구

Abstract

Control and measurement engineering technologies are applied to all industrial fields and being used conveniently. These technologies can reduce human error and allow us to work safely and maximize ease of use. In the case of automobiles, semi-autonomous driving technology has been developed for a while, allowing us to leave our hand. In recent years, autonomous driving is also in the commercialization stage in complex roads, front-collision prevention systems and highways. It has been a long time since this effort has been made to apply such control technology to marine vessels. Control techniques aimed at tracking trajectory routes in offshore marine environments have begun to be applied to ship systems since the early 1990s. The route tracking problem was the main point of view when the ship speed was low. This autonomous operation is possible at sea, thereby alleviating the fatigue of the sailors and improving the work environment. Unmanned ship operation technology capable of autonomous operation at sea was applied to small ships and special purpose ships (military object). As a result of continuous efforts for developing technologies is becoming a reality to apply unmanned technology such that the commercial vessel can maximize operational efficiency. For an example, in developed countries such as Norway and Finland, container test shop is being designed for practical use of unmanned technology, and a verification plan for actual sea areas is being established. In order to put the autonomous navigation and unmanned technology into practical use, it is necessary to measure and analyze the control technology by considering ship and marine environment conditions in real time. In order to operate the vessel safely, collision avoidance technology is necessary in addition to measurement technology, but an algorithm that identifies and corrects the physical characteristics of the ship in real time is also necessary. By fusing and optimizing other related technologies, it will be possible to operate autonomous ships more efficiently and safely. To become a fully autonomous vessel, automated ship berthing technology must be accompanied. A variety of studies have been carried out on the autonomous ship berthing. A multivariable neural controller design method for ship berthing has been suggested by using a multilayer feedforward neural network. This neural controller is designed to adjust parameters online to maintain robust control performance even under the effects of environmental disturbances. A typical example of research results about ship motion control relates to construction of a ship motion control system with four azimuth propellers. Although it is evaluated as the leading research result and utilized as DPS (Dynamic Positioning System) for special purpose ships such as FPSO and drill ship rather than general ship. The previous research results are not theoretical approaches but methods for attempting to connect the main propulsion unit and the rudder control only. In the previous researches, the target is to follow the established route under the environment conditions. Also, the change of the hydraulic force characteristics between the quay wall and the ship were not considered as approaching the quay wall. Autonomous vessels must perform all the processes from departure to berthing independently without external assistance. However, the berthing operation can not be accomplished only by the simple path following technique. Therefore, this paper aims to study and find out useful solution on the autonomous berthing problem with azimuth propeller and thrusters.