• Proceedings of the KIEE Conference
• /
• 2003.07a
• /
• pp.299-301
• /
• 2003

In this paper, a protection system using Multi-Agent concept for power distribution network is proposed. Multi agent system consist of Feeder agent, OCR(Over Current Relay) agent, Recloser agent and Switch agent. An agent calculates and corrects its parameter by itself through communication with neighboring agents and its own intelligence algorithm. Simulations in a simple distribution network show the effectiveness of the suggested protection system. Multi-Agent System, protection of distribution network, Communication.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.52 no.9
• /
• pp.506-512
• /
• 2003

In this paper, a communication module between Relay agents in a multi-agent system based power distribution network protection system is realized using DNP3.0(Distributed Network Protocol), which is the standard communication protocol of distribution automation system in KEPCO. The key words for agent communication in the multi-agent based protection system are defined and represented by use of DNP application function code. The communication module developed based on the proposed communication scheme is tested by use of the Communication Test Harness, a test tool for DNP protocol, then used to the multi-agent system based power distribution net work protection system.

• ETRI Journal
• /
• v.44 no.2
• /
• pp.194-207
• /
• 2022

In the last decade, substantial progress has been achieved in intelligent traffic control technologies to overcome consistent difficulties of traffic congestion and its adverse effect on smart cities. Edge computing is one such advanced progress facilitating real-time data transmission among vehicles and roadside units to mitigate congestion. An edge computing-based deep reinforcement learning system is demonstrated in this study that appropriately designs a multiobjective reward function for optimizing different objectives. The system seeks to overcome the challenge of evaluating actions with a simple numerical reward. The selection of reward functions has a significant impact on agents' ability to acquire the ideal behavior for managing multiple traffic signals in a large-scale road network. To ascertain effective reward functions, the agent is trained withusing the proximal policy optimization method in several deep neural network models, including the state-of-the-art transformer network. The system is verified using both hypothetical scenarios and real-world traffic maps. The comprehensive simulation outcomes demonstrate the potency of the suggested reward functions.

• Journal of Advanced Information Technology and Convergence
• /
• v.9 no.1
• /
• pp.127-134
• /
• 2019

In terms of deep Reinforcement Learning (RL), exploration can be worked stochastically in the action of a state space. On the other hands, exploitation can be done the proportion of well generalization behaviors. The balance of exploration and exploitation is extremely important for better results. The randomly selected action with ε-greedy for exploration has been regarded as a de facto method. There is an alternative method to add noise parameters into a neural network for richer exploration. However, it is not easy to predict or detect over-fitting with the stochastically exploration in the perturbed neural network. Moreover, the well-trained agents in RL do not necessarily prevent or detect over-fitting in the neural network. Therefore, we suggest a novel design of a deep RL by the balance of the exploration with drop-out to reduce over-fitting in the perturbed neural networks.

• Journal of Institute of Control, Robotics and Systems
• /
• v.18 no.7
• /
• pp.622-626
• /
• 2012

This paper presents synchronization of LTV (Linear Time-Varying) MAS (Multi-Agent Systems) with heterogeneous time-varying disturbances under a fixed, connected, and undirected communication network. All the agents can collect only relative state information from their neighborhoods. To achieve synchronization of the MAS, an integral control scheme is proposed based on relative state information between agents.

• Proceedings of the Korea Database Society Conference
• /
• 1999.06a
• /
• pp.125-133
• /
• 1999

The objective of knowledge discovery and data mining lies in the generation of useful insights from a store of data. This paper presents a framework for knowledge mining to provide a systematic approach to the selection and deployment of tools for automated learning. Every methodology has its strengths and limitations. Consequently, a multistrategy approach may be required to take advantage of the strengths of disparate technique while circumventing their individual limitations. For concreteness, the general framework for data mining in marketing is examined in the context of developing agents for optimizing a supply chain network.

• IMMUNE NETWORK
• /
• v.5 no.2
• /
• pp.105-109
• /
• 2005

Background: Dendritic cells (DCs) are the most potent APCs (antigen-presenting cells) and playa critical role in immune responses. Galectin-3 is a biological lectin with a beta-galactoside binding affinity. Recently, proteomic analysis revealed the presence of galectin-3 in the exosome of mature DCs. However, the expression and function of galectin-3 in DCs remains unclear yet. Methods: We used bone marrow-derived DCs of mouse and showed the expression of galectin-3 in DCs by using flow cytometry analysis and Western blot analysis. Results: Galectin-3 was determined as single band of 35 kDa in Western blot analysis. Flow cytometry analysis showed the major growth factor for DCs, granulocyte-macrophage colony stimulating factor (GM-CSF) and maturing agents, anti-CD40 monoclonal antibody (mAb) and lipopolysaccharide (LPS) consistently increased the intracellular expression of galectin-3 in DCs compared to medium alone. In addition, DCs treated with maturing agents did marginally express galectin-3 on their surface. Conclusion: This study suggests that galectin-3 in DCs may be regulated by critical factors for DC function.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 1999.03a
• /
• pp.125-133
• /
• 1999

The objective of knowledge discovery and data mining lies in the generation of useful insights from a store of data. This paper presents a framework for knowledge mining to provide a systematic approach to the selection and deployment of tools for automated learning. Every methodology has its strengths and limitations. Consequently, a multistrategy approach may be required to take advantage of the strengths of disparate technique while circumventing their individual limitations. For concreteness, the general framework for data mining in marketing is examined in the context of developing agents for optimizing a supply chain network.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.17 no.1
• /
• pp.39-45
• /
• 2007

In this paper, we developed an intelligent web based e-learning system based on multi agents. To do development of the system, we applied an inclination test that is based on the education theory to do grouping the desirable e-learning community. The proposed system, Intelligent Web based e-learning Multi Agent System (IMAS), is used the multi agents paradigm including learning manner by neural network for grouping of e-learning community and a new distributed multi agent framework proposed here.

• Proceedings of the IEEK Conference
• /
• 2009.05a
• /
• pp.91-93
• /
• 2009

In this paper, we present the Q-learning method for adaptive traffic signal control on the basis of In this paper, we present the Q-learning method for adaptive traffic signal control on the basis of multi-agent technology. The structure is composed of sixphase agents and one intersection agent. Wireless communication network provides the possibility of the cooperation of agents. As one kind of reinforcement learning, Q-learning is adopted as the algorithm of the control mechanism, which can acquire optical control strategies from delayed reward; furthermore, we adopt dynamic learning method instead of static method, which is more practical. Simulation result indicates that it is more effective than traditional signal system.