• 한국시뮬레이션학회:학술대회논문집
• /
• 한국시뮬레이션학회 2001년도 The Seoul International Simulation Conference
• /
• pp.321-324
• /
• 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 Beaming is the machine learning in which an agent learns from indirect, delayed reward an optimal policy to choose sequences of actions that produce the greatest cumulative reward. Therefore the reinforcement loaming is different from supervised teaming in the sense that there is no presentation of input-output pairs as training examples. Furthermore, model-free reinforcement loaming algorithms like Q-learning do not require defining or loaming 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 space 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)
• /
• 제7권10호
• /
• pp.2376-2394
• /
• 2013

In order to ensure both of the whole system capacity and users QoS requirements in heterogeneous wireless networks, admission control mechanism should be well designed. In this paper, Multi-agent Q-learning based Admission Control Mechanism (MQACM) is proposed to handle new and handoff call access problems appropriately. MQACM obtains the optimal decision policy by using an improved form of single-agent Q-learning method, Multi-agent Q-learning (MQ) method. MQ method is creatively introduced to solve the admission control problem in heterogeneous wireless networks in this paper. In addition, different priorities are allocated to multiple services aiming to make MQACM perform even well in congested network scenarios. It can be observed from both analysis and simulation results that our proposed method not only outperforms existing schemes with enhanced call blocking probability and handoff dropping probability performance, but also has better network universality and stability than other schemes.

• 인터넷정보학회논문지
• /
• 제5권5호
• /
• pp.69-77
• /
• 2004

인터넷이 널리 보급되면서 지능형 검색 에이전트들이 사용자의 요구를 만족시키기 위해 일반화되어 사용되고 있다. 그러나 이러한 지능형 멀티에이전트들은 서로 독립적으로 사용되어 멀티에이전트들 간의 분산된 정보를 원활하고 효율적으로 처리하기 위한 상호 협력 작용이 부족해 정보의 신뢰성이 낮고 동적으로 변화하는 분산 환경에 대처하기가 어렵다. 이런 문제를 해결하기 위해 본 논문에서는 멀티에이전트간의 효율적인 상호 협력과 빠른 정보처리를 위해 브로커 에이전트에 에이전시를 생성하고 신경망을 이용해 멀티에이전트들의 에이전시들을 분류하여 더욱 신속·정확한 정보를 사용자에게 제공하도록 한다. 또한 정보의 신뢰성을 위해서 에이전트 관리기법을 제안하여 기존의 검색 시스템이 가지고 있는 정보갱신문제를 향상시키고, 시뮬레이션을 통해 본 연구의 성능을 평가한다.

• 한국항공우주학회지
• /
• 제48권8호
• /
• pp.573-580
• /
• 2020

다중개체를 제어하기 위해서 해결해야 되는 문제들 중 하나는 위치제어다. 위치와 속도를 제어하기 위한 모델로 augmented Cucker-Smale 모델이 존재했다. 하지만 기존 모델은 모든 개체에 동일한 시스템을 적용함에 따라서 개별개체의 특성을 살리지 못했다는 특징이 있다. 본 논문에서는 그 점을 보안하고 적절한 형태로 변형하기 위해서 초기 위치와 분포를 이용한 마할라노비스 거리를 계수와 통계학적 자유도를 적용해서, 모델의 수렴시간과 소모에너지를 동시에 줄이고자 한다. 모델의 성능 검증을 위해서 몬테카를로 시뮬레이션을 통해서 전체적인 경향성을 판단했고, 추가적으로 개별 개체의 움직임을 분석하여서 마할라노비스 거리 계수가 적절한 역할을 수행하고 있는지 확인했다.

• 제어로봇시스템학회논문지
• /
• 제20권5호
• /
• pp.526-532
• /
• 2014

In this paper, we present and analyze a LQ (Linear Quadratic) inverse optimal state-consensus protocol for continuous-time multi-agent systems with undirected graph topology. By Lyapunov analysis of the state-consensus error dynamics, we show the sufficient conditions on the algebraic connectivity of the graph to guarantee LQ inverse optimality and closed-loop stability. A more relaxed stability condition is also provided in terms of the algebraic connectivity. Finally, a formation control protocol for multiple mobile robots is proposed based on the target LQ inverse optimal consensus protocol, and the simulation results are provided to verify the performance of the proposed LQ inverse formation control method.

• 제어로봇시스템학회논문지
• /
• 제20권12호
• /
• pp.1231-1237
• /
• 2014

본 논문에서는 축소 차수 외란 관측기를 이용하여 외란이 있는 이종 다개체 시스템을 위한 협조 추종 제어기를 제안하였다. 이를 위해 우선 주어진 제어 문제가 외란과 모델 불확실성을 가지는 시스템을 위한 강인 제어 문제로 변환 될 수 있음을 보이고 변환된 문제에 외란 관측기 기반의 동적 협조 추종 제어기를 설계하였다. 모의 실험을 통해서 제안하는 이종 다개체 시스템의 협조 추종을 성공적으로 달성함을 보였다.

• Transactions on Control, Automation and Systems Engineering
• /
• 제1권2호
• /
• pp.147-152
• /
• 1999

It is an interesting area in the field of artifical intelligence to find an analytic model of cooperative structure for multiagent system accomplishing a given task. Usually it is difficult to design controllers for multi-agent systems without a comprehensive knowledge about the system. One of the way to overcome this limitation is to implement an evolutionary approach to design the controllers. This paper introduces the use of a genetic algorithm to discover a fuzzy logic controller with rules that govern emergent agents solving a pursuit problem in a continuous world. Simulation results indicate that, given the complexity of the problem, an evolutionary approach to find the fuzzy logic controller seems to be promising.

• 제어로봇시스템학회논문지
• /
• 제18권5호
• /
• pp.439-444
• /
• 2012

In this study, we address the formation control problem of coupled unicycle-type mobile robots, each of which can interact with its neighboring robots by communicating their position outputs. Each communication link between two mobile robots is assumed to be established according to the given time-varying interconnection topology that switches within a finite set of connected fixed undirected networks and has a non-vanishing dwell time. Under this setup, we propose a distributed formation control algorithm by using the dynamics extension and feedback linearization methods, and by employing a consensus algorithm for linear multi-agent systems which provides arbitrary fast convergence rate to the agreement of the multi-agent system. Finally, the proposed result is demonstrated through a computer simulation.

• 제어로봇시스템학회논문지
• /
• 제20권12호
• /
• pp.1225-1230
• /
• 2014

This paper investigates the group average-consensus and group formation-consensus problems for first-order multi-agent systems. The control protocol for group consensus is designed by considering the positive adjacency elements. Since each intra-group Laplacian matrix cannot be satisfied with the in-degree balance because of the positive adjacency elements between groups, we decompose the Laplacian matrix into an intra-group Laplacian matrix and an inter-group Laplacian matrix. Moreover, average matrices are used in the control protocol to analyze the stability of multi-agent systems with a fixed and undirected communication topology. Using the graph theory and the Lyapunov functional, stability analysis is performed for group average-consensus and group formation-consensus, respectively. Finally, some simulation results are presented to validate the effectiveness of the proposed control protocol for group consensus.

• 한국산학기술학회논문지
• /
• 제15권12호
• /
• pp.7451-7458
• /
• 2014

본 논문에서는 독립된 풍력 복합발전 시스템을 대상으로 시스템의 모델링과 다양한 환경에서 체계적으로 동작시키기 위한 멀티에이전트 기반의 제어방법을 제안한다. 멀티에이전트 제어는 풍력발전기, 디젤발전기, 배터리, 부하로 구성되는 새로운 형식의 하이브리드 제어방법이고, 풍속과 배터리의 충전상태에 따라 풍력 복합발전 시스템의 운전은 14개의 모드로 나누어 수행된다. 시뮬레이션 성능평가를 통해 제안된 알고리즘이 독립된 풍력 복합발전 시스템에서 다양한 풍속변화가 존재하는 경우에도 효율적으로 운전될 수 있음을 보여준다.