• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.1
• /
• pp.42-47
• /
• 2015

This paper presents a rank-based formation for multiple agents based on potential functions, where the proposed method uses the relative position of two neighboring agents. The conventional formation scheme of multiple systems requires communication between agents and a central computer to get the positions of all multiple agents. In the study, differently from previous studies, the formation scheme uses the relative position of two neighboring agents in a local coordinate system. In addition, it introduces a singular agent association that considers only the relative position between an agent and its neighboring agents, instead of multiple associations among all information about all agents. Furthermore, the proposed framework explores the benefits of different formation types. Extensive simulation results show that the proposed approach verifies the viability and effectiveness of the proposed formation.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.15 no.7
• /
• pp.834-839
• /
• 2005

When a great variety of services become available to user through the broadband convergence network in the ubiquitous home environment, an intelligent agent is required to deal with the complexity of services and perceive intension of a user. Different from the old-fashioned command-based user interface for selecting services, conversation enables flexible and rich interactions between human and agents, but diverse expressions of the user's background and context make conversation hard to implement by using either user-initiative or system-initiative methods. To deal with the ambiguity of diverse expressions between user and agents, we have to apply hierarchial bayesian networks for the mixed initiative conversation. Missing information from user's query is analyzed by hierarchial bayesian networks to inference the user's intension so that can be collected through the agent's query. We have implemented this approach in ubiquitous home environment by implementing simulation program.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.8 no.5
• /
• pp.151-159
• /
• 2009

Network models for pedestrian flows have been studied in various ways. However, because of the simplicity and application, a number of researchers prefer the CA Model to analyze pedestrian's complicated behavior. These kinds of models based on Agent are being used as a microscopic analyzing method since it can easily adapt individuals' various characters and movement types. However, because pedestrians' movement can be (easily) effected by where they are and where they head, some models using the same rules have limit when considering pedestrians' every different movement. In this research, homogeneous section is defined as a similar movement type of individuals. With MDPM, we suggest simulation method explaining one-way walk and side-walk which could not be done in past.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.21 no.6
• /
• pp.712-717
• /
• 2011

Recently, diverse schedule management agents are being researched for the efficient schedule management of smart devices users, but they remain at a confirmatory level. In order to efficiently manage user's schedules, execution of planned schedules should be monitored to help users properly execute their schedules, or feedback must be given so that when setting up new schedules, users can plan their schedule according to their schedule establishment patterns. This research proposes a schedule management agent that infers the user's behaviors by using acquired user context, and provides schedule related feedback depending on the user's behavior patterns, when users are executing their schedules or planning new schedules. For this, collected user context information is preprocessed and user's behavior is inferred by Bayesian network. Also, in order to provide feedbacks necessary for confirming the user's schedule execution and new schedule establishment, a context tree pattern matching method for the user's schedule, location and time contexts was applied, then verified with 6 weeks of user simulation in a mobile environment.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2023.05a
• /
• pp.132-133
• /
• 2023

The safe evacuation of people during disasters is of utmost importance. Various life safety evacuation simulation tools have been developed and implemented, with most relying on algorithms that analyze maps to extract the shortest path and guide agents along predetermined routes. While effective in predicting evacuation routes in stable disaster conditions and short timeframes, this approach falls short in dynamic situations where disaster scenarios constantly change. Existing algorithms struggle to respond to such scenarios, prompting the need for a more adaptive evacuation route algorithm that can respond to changing disasters. Artificial intelligence technology based on reinforcement learning holds the potential to develop such an algorithm. As a fundamental step in algorithm development, this study aims to evaluate whether an evacuation algorithm developed by reinforcement learning satisfies the performance conditions of the evacuation simulation tool required by IMO MSC.1/Circ1533.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.8B
• /
• pp.1105-1114
• /
• 2010

The Proxy Mobile IPv6 (PMIPv6) has been standardized by the IETF NETLMM WG for network-based mobility management. The PMIPv6 can provide IP mobility for Mobile Nodes (MNs) with low handover latency and less wireless resource usage. But, since the PMIPv6 is basically designed for local mobility management, it cannot support directly global mobility management between different PMIPv6 domains. In the PMIPv6, since all traffic is routed through a Local Mobility Anchor (LMA), it causes a long end-to-end delay and triangular routing problem. Therefore, in this paper, we propose a fast network-based global mobility management scheme and route optimization scheme with a new network entity, called Global Mobility Agent (GMA). Numerical analysis and simulation results show that the proposed scheme is able to support global mobility between different public domains with low handover latency and low end-to-end delay, compared with the PMIPv6.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.21 no.2
• /
• pp.94-106
• /
• 2018

The epidemiological model is useful for creating simulation and associated preventive measures for disease spread, and provides a detailed understanding of the spread of disease space through contact with individuals. In this study, propose an agent-based spatial model(ABM) integrated with spatial big data to simulate the spread of MERS-CoV infections in real time as a result of the interaction between individuals in space. The model described direct contact between individuals and hospitals, taking into account three factors : population, time, and space. The dynamic relationship of the population was based on the MERS-CoV case in Seoul Metropolitan Government in 2015. The model was used to predict the occurrence of MERS, compare the actual spread of MERS with the results of this model by time series, and verify the validity of the model by applying various scenarios. Testing various preventive measures using the measures proposed to select a quarantine strategy in the event of MERS-CoV outbreaks is expected to play an important role in controlling the spread of MERS-CoV.

• Journal of Korea Game Society
• /
• v.16 no.2
• /
• pp.27-40
• /
• 2016

Predicting stealthy behaviors plays an important role in designing stealth games. It is, however, difficult to automate this task because human players interact with dynamic environments in real time. In this paper, we present a reinforcement learning (RL) method for simulating stealthy movements in dynamic environments, in which an integrated model of Q-learning with Artificial Neural Networks (ANN) is exploited as an action classifier. Experiment results show that our simulation agent responds sensitively to dynamic situations and thus is useful for game level designer to determine various parameters for game.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.19 no.3
• /
• pp.131-139
• /
• 2024

Crew Transfer Vessels (CTVs) are primarily used for the maintenance of offshore wind farms. Despite being manually operated by professional captains and crew, collisions with other ships and marine structures still occur. To prevent this, the introduction of autonomous navigation systems to CTVs is necessary. In this study, research on the obstacle avoidance system of the autonomous navigation system for CTVs was conducted. In particular, research on obstacle avoidance simulation for CTVs using deep reinforcement learning was carried out, taking into account the currents and wind loads in offshore wind farms. For this purpose, 3 degrees of freedom ship maneuvering modeling for CTVs considering the currents and wind loads in offshore wind farms was performed, and a simulation environment for offshore wind farms was implemented to train and test the deep reinforcement learning agent. Specifically, this study conducted research on obstacle avoidance maneuvers using MATD3 within deep reinforcement learning, and as a result, it was confirmed that the model, which underwent training over 10,000 episodes, could successfully avoid both static and moving obstacles. This confirms the conclusion that the application of the methods proposed in this study can successfully facilitate obstacle avoidance for autonomous navigation CTVs within offshore wind farms.

• The KIPS Transactions:PartC
• /
• v.16C no.6
• /
• pp.707-716
• /
• 2009

Wireless Sensor Networks are consisted of sensor nodes that communicated with each other to transmit information. Because sensor nodes have physically many limits, wireless sensor networks are hard to adopt for traditional networks. Transmissions are consumed most energy of sensor nodes. That's why energy-efficient transmission techniques and QoS support techniques for different kind of data are most important in wireless sensor networks. The thesis proposes the agent based framework for energy distribution and QoS in wireless sensor networks. Agents have its own behavior policy by means of a gene, which is optimized by genetic operations. Agents behavior to distribute energy consumption over sensor nodes. Simulation results show that the enhanced framework extends the lifetime of sensor nodes. Successful transmission ratios of emergency data and non emergency data are increased by 27% and 14%, respectively. Also, the results demonstrate that Qos of networks are improved.