• 한국항해항만학회지
• /
• 제29권4호
• /
• pp.299-304
• /
• 2005

A simulation system is needed to train students and mariners in order that they can take suitable actions to evade typhoon's strike promptly and sufficiently. In order to make such kind of system, three kinds of models about the typhoon are necessary, typhoon prediction model to generate typhoon's track, wind & wave-field model to make sea conditions around the typhoon and evaluation model of trainee's action whether their actions were suitable or not during simulation. We have developed the prediction and wind & wave-field models of typhoon, but the evaluation model has not been developed yet. In this paper, after making a method for evaluating trainee's actions by seakeeping performance, we propose an typhoon avoidance simulation system for training mariners so that they can promote their abilities to evade the typhoons at sea.

• Journal of Information Processing Systems
• /
• 제8권3호
• /
• pp.409-420
• /
• 2012

The decision-making by agents in games is commonly based on reinforcement learning. To improve the quality of agents, it is necessary to solve the problems of the time and state space that are required for learning. Such problems can be solved by Macro-Actions, which are defined and executed by a sequence of primitive actions. In this line of research, the learning time is reduced by cutting down the number of policy decisions by agents. Macro-Actions were originally defined as combinations of the same primitive actions. Based on studies that showed the generation of Macro-Actions by learning, Macro-Actions are now thought to consist of diverse kinds of primitive actions. However an enormous amount of learning time and state space are required to generate Macro-Actions. To resolve these issues, we can apply insights from studies on the learning of tasks through Programming by Demonstration (PbD) to generate Macro-Actions that reduce the learning time and state space. In this paper, we propose a method to define and execute Macro-Actions. Macro-Actions are learned from a human subject via PbD and a policy is learned by reinforcement learning. In an experiment, the proposed method was applied to a car simulation to verify the scalability of the proposed method. Data was collected from the driving control of a human subject, and then the Macro-Actions that are required for running a car were generated. Furthermore, the policy that is necessary for driving on a track was learned. The acquisition of Macro-Actions by PbD reduced the driving time by about 16% compared to the case in which Macro-Actions were directly defined by a human subject. In addition, the learning time was also reduced by a faster convergence of the optimum policies.

• 한국항해항만학회:학술대회논문집
• /
• 한국항해항만학회 2007년도 추계학술대회 및 제23회 정기총회
• /
• pp.79-80
• /
• 2007

An expert system of collision avoidance developed by CLIPS and Visual C++ is continuously introduced in this paper. Further, a simulation function of collision avoidance is added to the expert system, the function can simulate the avoidance actions of own ship and a specific target of a period of future time. This function can help navigators to estimate collision risk and make proper collision avoidance actions in dangerous situations for navigational safety of ships. Furthermore, navigational rules can also be evaluated during the process of simulation.

• 한국게임학회 논문지
• /
• 제8권1호
• /
• pp.79-90
• /
• 2008

본 논문에서는 동기 기반의 계층적 행동 계획 시스템을 제안한다. 가상 시뮬레이션 게임 환경에서 에이전트는 행동 계획 시스템을 통해 적합한 행동을 선택하게 된다. 행동 선택 시스템은 동기를 추출하고 목표를 선택하고 행동을 생성하고 최적화를 수행한다. 동기를 평가할 때 갑작스럽게 발생하거나 누적된 이벤트에 대해 계산한다. 동기를 선택할 때는 확률 분포를 사용하여 무작위로 선택한다. 계층적 목표 트리를 탐색한 후에 목표를 실행할 수 있다. 행동들을 비교한 후 가장 적합한 행동을 선택하게 된다. 선택을 할 때 안전도 값과 만족도 값을 비교하여 최적화된 행동을 선택한다. 본 연구에서 제안한 시스템을 식당경영 게임에 적용했다.

• International Journal of Contents
• /
• 제15권4호
• /
• pp.16-26
• /
• 2019

The efficient implementation of various physical actions of agents to respond to dynamically changing situations is essential for the simulation of realistic agents and activities in a cyber world. To achieve a maximum diversity of actions and immediate responsiveness to abrupt changes in situations, we have developed an animation technique in which complex actions are recursively constructed by reusing a set of primitive motions, and agents are designed to react in real-time to abrupt ambient changes by computationally satisfying kinematic constraints on body parts with respect to their goals. Our reusing scheme is extended to visualize the procedure of realistic intricate situations involving many concurring events. Our approach based on motion reuse and recursive assembly has clear advantages in motion variability and action diversity with respect to authoring scalability and motion responsiveness compared to conventional monolithic (static) animation techniques. This diversity also serves to accommodate the characteristic unpredictability of events concurring in a situation due to inherent non-determinism of associated conditions. To demonstrate the viability of our approach, we implement several composite and parallel actions in a dynamically changing example situation involving events that were originally independent until coincidentally inter-coupled therein.

• 제어로봇시스템학회논문지
• /
• 제8권8호
• /
• pp.665-672
• /
• 2002

This paper deals with the human-interactive actions of a quadruped robot by using Genetic Algorithm. In case we have to work out the designed plan under the special environments, our robot will be required to have walking capability, and patterns with legs, which are designed like gaits of insect, dog and human. Our quadruped robot (called SERO) is capable of not only the basic actions operated with sensors and actuators but also the various advanced actions including walking trajectories, which are generated by Genetic Algorithm. In this paper, the body and the controller structures are proposed and kinematics analysis are performed. All of the suggested motions of SERO are generated by PC simulation and implemented in real environment successfully.

• Nuclear Engineering and Technology
• /
• 제53권3호
• /
• pp.825-832
• /
• 2021

A simulation-based approach is proposed to study the protective actions taken by residents during nuclear emergencies using cognitive findings. Human perception-based behaviors are not heavily incorporated in the evacuation study for nuclear emergencies despite their known importance. This study proposes a generic framework of perception-based behavior simulation, in accordance with the ecological concept of affordance theory and a formal representation of affordance-based finite state automata. Based on the generic framework, a simulation model is developed to allow an evacuee to perceive available actions and execute one of them according to Newton's laws of motion. The case of a shadow evacuation under nuclear emergency is utilized to demonstrate the applicability of the proposed framework. The illustrated planning algorithm enables residents to compute not only prior knowledge of the environmental map, but also the perception of dynamic surroundings, using widely observed heuristics. The simulation results show that the temporal and spatial dynamics of the evacuation behaviors can be analyzed based on individual perception of circumstances, while utilizing the findings in cognitive science under unavoidable data restriction of nuclear emergencies. The perception-based analysis of the proposed framework is expected to enhance nuclear safety technology by complementing macroscopic analyses for advanced protective measures.

• 한국해양공학회지
• /
• 제36권1호
• /
• pp.41-49
• /
• 2022

With the advancement of intelligent vehicles and unmanned systems, there is a growing interest in underwater surveys using autonomous marine vehicles (AMVs). This study presents an automated planning strategy for a long-term survey mission using a fleet of AMVs consisting of autonomous surface vehicles and autonomous underwater vehicles. Due to the complex nature of the mission, the actions of the vehicle must be of high-level abstraction, which means that the actions indicate not only motion of the vehicle but also symbols and semantics, such as those corresponding to deploy, charge, and survey. For automated planning, the planning domain definition language (PDDL) was employed to construct a mission planner for realizing a powerful and flexible planning system. Despite being able to handle abstract actions, such high-level planners have difficulty in efficiently optimizing numerical objectives such as obtaining the shortest route given multiple destinations. To alleviate this issue, a widely known technique in operations research was additionally employed, which limited the solution space so that the high-level planner could devise efficient plans. For a comprehensive evaluation of the proposed method, various PDDL-based planners with different parameter settings were implemented, and their performances were compared through simulation. The simulation result shows that the proposed method outperformed the baseline solutions by yielding plans that completed the missions more quickly, thereby demonstrating the efficacy of the proposed methodology.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 1997년도 한국자동제어학술회의논문집; 한국전력공사 서울연수원; 17-18 Oct. 1997
• /
• pp.648-651
• /
• 1997

This paper proposes methods for the cooperative control of multiple mobile robots and constructs a robotic soccer system in which the cooperation will be implemented as a pass play of two robots. To play a soccer game, elementary actions such as shooting and moving have been designed, and Q-learning, which is one of the popular methods for reinforcement learning, is used to determine what actions to take. Through simulation, learning is successful in case of deliberate initial arrangements of ball and robots, thereby cooperative work can be accomplished.

• 국제학술발표논문집
• /
• The 3th International Conference on Construction Engineering and Project Management
• /
• pp.732-737
• /
• 2009

Due to various project delivery methods and the complexity of construction projects in the construction industry, developing the framework of construction management for critical, highly complex projects in the construction industry has become problematic. Currently, a set of construction manuals play a pivotal role in planning and managing construction projects as subcontractors try to complete their scope of work according to the instructions of a general contractor. It is challenging for general contractors to write a construction management procedure manual to cover various types of project delivery methods and construction projects. In construction, the construction procedure manuals describe specific actions to be taken through the project. In reality a few contactors own such manuals and their construction schedules include more construction operation activities. Thus, it is hard to estimate the workload and productivity of construction managers because the manual and the schedule do not present the amount of management efforts required to complete a project. This paper proposes a framework to present construction management tasks according to project delivery methods which can be applied to various construction projects. Actions for management tasks were mapped and were integrated with construction activities throughout the project life cycle. The framework can then be used to give specific instructions to project participants, collect management actions, and replicate management actions throughout the project life cycle. The framework can also be can used to visualize complete construction project to analyze and manage construction management activities in each phase of a project in order to enhance productivity and efficiency. The studies of existing construction manuals were carried out to identify construction managers' responsibilities. An artificial intelligence program, CLIPS (C-Language Integrated Production System) was used to search for appropriate actions for impending tasks from a set of predefined actions to be performed for a given situation. The framework would significantly help construction managers to understand interrelations among management tasks or actions within a project. Furthermore, the framework can be embedded into Building Information Modeling (BIM) or Facility Management Systems (FMS) so that designers and constructors would execute constructability review before construction begins.