In real system application, the interacting multiple model (IMM)-based uncertain target tracking system operates with the following problems: it requires less computing resources as well as a robust performance with respect to the maneuvering such as a sub-model mismatched case, and further, it requires an easy design procedure related to its structures and parameters. To solve these problems, an adaptive fuzzy IMM (AFIMM) algorithm, which is based on well-defined basis sub-models and well-adjusted mode transition probabilities (MTPs), is proposed. The basis sub-models are defined by the detailed analysis in terms of kinematic models as well as the maneuvering property and the MTPs are adjusted by the fuzzy adaptor as well as the fuzzy decision maker. To verify the performance of the proposed algorithm, an airborne target tracking is performed. Simulation results show that the AFIMM effectively solves the problems experienced in the uncertain target tracking system online.

서 론해양 시스템 개발로봇 하드웨어 제작 과정은 다음과 같다. 먼저, Fig. 1과

There is a gradual increase in the need for energy charging in marine environments because of energy limitations experienced by electric ships and marine robots. Buoys are considered potential energy charging systems, but there are several challenges, which include the need to maintain a fixed position and avoid hazards, dock with ships and robots in order to charge them, be robust to actions by birds, ships, and robots. To solve these problems, this study proposes a smart buoy robot that has multiple thrusters, multiple docking and charging parts, a bird spike, a radar reflector, a light, a camera, and an anchor, and its mechanism is developed. To verify the performance of the smart buoy robot, the position control under disturbance due to wave currents and functional tests such as docking, charging, lighting, and anchoring are performed. Experimental results show that the smart buoy robot can operate under disturbances and is functionally effective. Therefore, the smart buoy robot is suitable as an energy charging system and has potential in realistic applications.

No abstract

In recent years, the interest in oceans has been increasing due to the Fourth Industrial Revolution for Oceans, wind power, offshore photovoltaic power plants, and Antarctic and Arctic exploration bases.Accordingly, interest in the importance of maritime domain awareness (MDA) has also been increasing (Kim et al., 2016).Research on marine robots for efficient mission performance in oceans has increased with the increasing interest in MDA (Yeu et al., 2019, Park et al., 2021. In the case of autonomous marine robots, there is a problem of energy insufficiency because of the size limitation. The energy insufficiency leads to severe problems, such as loss or damage of the robots caused by the lack of energy when the robots are conducting highly difficult missions and the activity radius is limited (Park et al., 2019).To solve this problem, diverse studies have been conducted on marine-energy harvesting-based robots such as wave gliders and underwater gliders in South Korea and abroad. Among them, sailing-type robots, which have strengths in terms of speed and duration, stand out (Meinig et al., 2015).A sail drone, a sailing-type robot, is a system that is propelled by wind and is heavily affected by wind direction and speed. Hence, it has a disadvantage that the straight path movement is not excellent because of the cross-flow forces produced by wind (Sa et al., 2019).To solve this problem, research was conducted in South Korea on a type of sail drone, to which a keel, a cross-flow force prevention device, was attached; however, the manufacturing and maintenance costs were high because the attitude stability and modularization for maintenance and repair were not considered in the design and manufacturing process (Man et al., 2020). Furthermore, they did not consider suitable sailing control techniques for the sail drone.In this study, therefore, we developed a performance estimation tool that can ensure attitude stability using previously produced 3D design results and the smoothed-particle hydrodynamics (SPH) method, thereby securing attitude stability in the design process. Furthermore, we considered a V-model of the bow for speed improvement and modularization for robot maintenance and repair.We considered the shape of the catamaran-type sail drone (CSD) to increase the attitude stability and considered a hybrid propulsion method, in which the sail is used as the main propulsion method and an electric propeller is used as the auxiliary propulsion method.Recently, studies have used neural network (NN) methods (Fang et al., 2017) and deep learning (Sun and Gao, 2020) for the heading control of marine robots. However, owing to the complexity of sailing techniques, marine robots need expert systems that operate based on expert knowledge and are sensitive to wind direction and speed, thus