• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2000.11a
• /
• pp.383-386
• /
• 2000

웹상의 수많은 정보와 서비스들은 사용자의 업무 생산성과 의사결정의 질을 향상시키고 있다. 그러나 기존의 키워드 중심의 검색 엔진들은 질의에 대한 결과에 쓸모없는 정보들이 많아서 원하는 시간에 필요한 정보들을 찾는데 효율적이지 못한 문제점을 가지고 있었다. 따라서 본 논문은 개인의 정보를 에이전트가 자동적으로 관리하고, 개인의 선호도에 맞게 에이전트에게 검색 명령을 내리고, 사용자가 많이 접근하고 사용하는 정보를 자동적으로 관리하는 멀티 에이전트 시스템을 제시한다. 현재 구축중인 멀티 에이전트 시스템을 웹 상에 존재하는 수많은 정보 중에서 도서와 관련된 정보 검색을 대상으로 개발하고 있다.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2005.11a
• /
• pp.507-510
• /
• 2005

이 논문에서는 시맨틱 웹 서비스의 기술들을 이용하여 에이전트들이 상황을 과악하고 작업계획을 동적으로 수행할 수 있는 다중-에이전트 기반의 시맨틱 웹서비스 시스템을 설계하고 도서관 서비스 업무에 적용해 구현 하였다.

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.49.2-49
• /
• 2002

This paper describes the configuration of a multi-agent system that can recognize human intentions. This system constructs ontologies of human intentions and enables knowledge acquisition and sharing between intelligent agents operating in different environments. This is achieved by using a bi-directional associative memory network. The process of intention recognition is based on fuzzy association inferences. This paper shows the process of information sharing by using ontologies. The purpose of this research is to create human-centered systems that can provide a natural interface in their interaction with people.

• Proceedings of the Korea Society for Industrial Systems Conference
• /
• 2002.06a
• /
• pp.156-160
• /
• 2002

본 논문에서는 사용자의 속성과 물품에 대한 속성을 이용한 효과적인 경매시스템의 알고리즘에 대하여 연구하였다. 사람이 경매를 하는 것과 유사한 척도를 가지고 물품에 대한 가격을 결정하고 입찰가를 제안할 수 있는 알고리즘을 제안하였다. 이것은 본 연구에서는 물품에 대한 다속성을 이용하여 다수의 입찰자들의 제한된 자율성을 가진 agent들이 경매를 참여할 수 있도록 하였다. 또한 입찰자가 물품에 대한 금액을 산정 하는데 보다 객관적으로 판단을 할 수 있도록 보조하여준다. 이러한 이유들로 인하여 본 논문에서 제안한 알고리즘이 사람과 유사한 경매를 할 수 있을 뿐만 아니라, 사람의 감성부분도 속성에 추가할 수 있는 특징이 있다.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.16 no.12
• /
• pp.8908-8914
• /
• 2015

The Internet continues to grow and develop, but there are going to generate a variety of Internet attacks that exploit it. In the initial Internet environment, the attackers maliciously exploited Internet environments for ostentations and hobbies. but these days many malicious attempts purpose the financial gain so systematic and sophisticated attacks that are associated with various crimes are occurred. The structures, such as viruses and worms were present in the form of one source multi-target before. but recently, APT(Advanced Persistent Threat, intelligent continuous attacks) in the form of multi-source single target is dealing massive damage. The performance evaluation analyzed whether to generate audit data and detect integrity infringement, and false positives for normal traffic, process detecting and blocking functions, and Agent policy capabilities with respect to the application availability.

• Journal of Korean Society of Transportation
• /
• v.21 no.1
• /
• pp.147-156
• /
• 2003

교통문제는 사회시스템이 복잡해짐에 따라 더욱 대처하기 어려운 국면으로 가고 있다. 따라서 교통시스템의 변화 예측도 용이하지 않다. 복잡계(complex system)의 하나로 볼 수 있는 교통시스템을 대처하는데 있어 전통적인 상의하달(上意下達) 접근에 한계가 있음을 부인할 수 없다. 지난 10여년 동안 물리학, 시스템공학, 컴퓨터공학 분야 등 다양한 분야에서 복잡계에 대한 활발한 연구가 진행되고 있다. 복잡계의 해석을 위한 새로운 개념과 접근기법들이 도입되고 있는데, 그들 중 에이젠트기반모형(agent-based modelling)은 교통분야에 적용 가능한 매우 흥미를 있는 기법으로 보인다. 본 글에서는 에이젠트기반모형이 무엇이며, 어떻게 사용되고 있으며, 교통분야에서의 적용가능성을 검토하였다. 본 글에서 제시한 에이젠트기반접근은 기존 방법과는 다른 하의상달(下意上達) 방식의 기법이다. 이는 시스템의 개별 구성원인 에이젠트의 행태와 에이젠트간의 상호작응에 초점을 둔다. 에이젠트의 행태와 상호작용의 규칙이 변함에 따라 전 에이젠트시스템에 나타나는 변화를 추적할 수 있다. 오늘날 교통문제의 복잡성은 교통시스템의 더욱 세분화된 하부시스템의 다양화와 상호작용, 그리고 개별 차량 또는 운전자의 행태와 상호작용에서 기인된다고 볼 수 있다. 따라서 에이젠트기반의 접근은 아직 연구는 미흡하지만 복잡한 교통시스템의 운영과 분석에 적용잠재력이 큰 기법으로 판단된다.

• Asia pacific journal of information systems
• /
• v.12 no.2
• /
• pp.197-216
• /
• 2002

Cooperativeness within an organization can be conceptualized as the degree of members' willingness to work with others. The simulation study investigates the relationships of cooperativeness with team performance at different levels of information redundancy by using a multi-agents model called Team-Soar. The model consists of a group of four individual Al agents situated in a network, which models a naval command and control team consisting of four members. The study used a $9{\times}3$ design in which agent cooperativeness was manipulated at nine levels by gradually replacing selfish team members with increasing numbers of neutral and cooperative members, while information redundancy was controlled at three different levels(i.e., low, medium, and high). Results of the Team-Soar simulation show that cooperation has positive impacts on team performance. Further, the results reveal that the impact of agent cooperativeness on team performance depends on the amount of information needed to be processed during the decision making process.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.50 no.4
• /
• pp.188-196
• /
• 2001

This paper presents and efficient algorithm for the loss minimization by automatic sectionalizing switch operation in distribution systems. Ant colony algorithm is multi-agent system in which the behaviour of each single agent, called artificial ant, is inspired by the behaviour of real ants. Ant colony algorithm is suitable for combinatiorial optimization problem as network reconfiguration because it use the long term memory, called pheromone, and heuristic information with the property of the problem. The proposed methodology with some adoptions have been applied to improve the computation time and convergence property. Numerical examples demonstrate the validity and effectiveness of the proposed methodology using a KEPCO's distribution system.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2001.01a
• /
• pp.59-64
• /
• 2001

The robot soccer simulation game is a dynamic multi-agent environment. In this paper we suggest a new reinforcement learning approach to each agent's dynamic positioning in such dynamic environment. Reinforcement learning is the machine learning in which an agent learns from indirect, delayed reward an optimal policy to chose sequences of actions that produce the greatest cumulative reward. Therefore the reinforcement learning is different from supervised learning in the sense that there is no presentation of input pairs as training examples. Furthermore, model-free reinforcement learning algorithms like Q-learning do not require defining or learning any models of the surrounding environment. Nevertheless it can learn the optimal policy if the agent can visit every state- action pair infinitely. However, the biggest problem of monolithic reinforcement learning is that its straightforward applications do not successfully scale up to more complex environments due to the intractable large space of states. In order to address this problem. we suggest Adaptive Mediation-based Modular Q-Learning (AMMQL)as an improvement of the existing Modular Q-Learning (MQL). While simple modular Q-learning combines the results from each learning module in a fixed way, AMMQL combines them in a more flexible way by assigning different weight to each module according to its contribution to rewards. Therefore in addition to resolving the problem of large state effectively, AMMQL can show higher adaptability to environmental changes than pure MQL. This paper introduces the concept of AMMQL and presents details of its application into dynamic positioning of robot soccer agents.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.7
• /
• pp.3566-3582
• /
• 2019

In intelligent warehouse picking system, the allocation of tasks has an important influence on the efficiency of the whole system because of the large number of robots and orders. The paper proposes a method to solve the task allocation problem that multi-robot task allocation problem is transformed into transportation problem to find a collision-free task allocation scheme and then improve the capability of task processing. The task time window and the power consumption of multi-robot (driving distance) are regarded as the utility function and the maximized utility function is the objective function. Then an integer programming formulation is constructed considering the number of task assignment on an agent according to their battery consumption restriction. The problem of task allocation is solved by table working method. Finally, simulation modeling of the methods based on table working method is carried out. Results show that the method has good performance and can improve the efficiency of the task execution.