• Journal of Food Hygiene and Safety
• /
• v.17 no.3
• /
• pp.137-145
• /
• 2002

Tributyltin compounds have been increasingly used in the form of plastic stabilizers, catalytic agents, industrial agricultural biocides, antifouling paint, and pesticides. Among these organotin compounds, large amounts of tributyltin(TBT) and triphenyltin(TPT) have been used as antifouling agents because they have a superior ability to prevent marine organism from being encrusted on ship bottoms and in culturing nets. Environmental pollution by these organotin compounds in the aquatic environment were undertaken. The international maritime Organization's established a provisional tolerable daily intake(TDI) of 1.6[micro]g TBTO/kg/ B.W. The Food and Agiculture Organization (of the United Nations)/world Health Organization's (FAO/WHO) proposed a TDI of 0.5ug TPT/kg BW/d. This study is conducted monitoring of TBT on seafoods in Korea and risk assessment for exposure on TBT in seafoods. Total hazard index(using Reference Dose : 0.3 ug TBTO/kg B.W/day) of intake exposure on seafoods is 0.04 as the 50th percentile, 0.08 as the 95th percentile. This value is estimated by Monte-Carlo simulation using Crystal Ball(Decisioneering Co., 2001).

• Journal of Korea Multimedia Society
• /
• v.9 no.5
• /
• pp.649-657
• /
• 2006

Reinforcement Learning is a computational approach to learning whereby an agent take an action which maximize the total amount of reward it receives among possible actions within current state when interacting with a uncertain environment. Q-learning, one of the most active algorithm in Reinforcement Learning, is consist of rewards which is obtained when an agent take an action. But it has the problem with mapping real world to discrete states. When state spaces are very large, Q-learning suffers from time for learning. In constant, when the state space is reduced, many state spaces map to single state space. Because an agent only learns single action within many states, an agent takes an action monotonously. In this paper, to reduce time for learning and complement simple action, we propose the Q-learning using influence map(QIM). By using influence map and adjacent state space's learning result, an agent could choose proper action within uncertain state where an agent does not learn. When this paper compares simulation results of QIM and Q-learning, we show that QIM effects as same as Q-learning even thought QIM uses 4.6% of the Q-learning's state spaces. This is because QIM learns faster than Q-learning about 2.77 times and the state spaces which is needed to learn is reduced, so the occurred problem is complemented by the influence map.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.22 no.6
• /
• pp.722-728
• /
• 2012

Urban sewer systems has a limitation of capacity of rainwater storage and problem of occurrence of untreated sewage, so adopting a storage facility for sewer flooding prevention and urban non-point pollution reduction has a big attention. The Korea Ministry of Environment has recently introduced a new concept of "multi-functional storage facility", which is crucial not only in preventive stormwater management but also in dealing with combined sewer overflow and sanitary sewer discharge, and also has been promoting its adoption. However, reserving a space for a single large-scale storage facility might be difficult especially in urban areas. Thus, decentralized construction of small- and midium-sized storage facilities and its operation have been introduced as an alternative way. In this paper, we propose a model predictive control scheme for an optimized operation of distributed storage facilities and sewer networks. To this aim, we first describe the mathematical model of each component of networks system which enables us to analyze its detailed dynamic behavior. Second, overflow locations and volumes will be predicted based on the developed network model with data on the external inflow occurred at specific locations of the network. MPC scheme based on the introduced particle swarm optimization technique then produces the optimized the gate setting for sewer network flow control, which minimizes sewer flooding and maximizes the potential storage capacity. Finally, the operational efficacy of the proposed control scheme is demonstrated by simulation study with virtual rainstorm event.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.12 no.3
• /
• pp.45-55
• /
• 2009

This study was intended to increase efficiency of traffic flow management on intersection. The result suggested to establish a left-turn at own risk lane to increase efficiency of traffic flow on intersection. The scope of the research was to investigate the geometric structure of a signal-controlled intersection, traffic volume(density) with respect to directions and traffic signal display, and to select a signalling intersection into which a car waiting for a traffic signal enters by adjusting the display sequence of traffic signal. The delay with respect to directions and for the whole intersection was compared for the current situation and an improvement plan. Using TSIS, a traffic analysis package, the traffic situation on an intersection was investigated. Based on the simulation result for Seok-Jeon intersection in Ma-San selected from the field investigation of intersections to which an improvement plans would be applicable, the waiting time in the direction without a entering traffic signal was decreased to be 78.6 seconds per car and that of the direction expecting the increase of waiting time was increased by 4 seconds per car only. It was confirmed that the waiting time for the whole intersection was improved.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.47 no.2
• /
• pp.103-112
• /
• 2010

In sensor networks, because sensors are deployed in an unprotected environment, they are prone to be targets of compromise attack, If the number of compromised nodes increases considerably, the key management in the network is paralyzed. In particular, compromise of Cluster Heads (CHs) in clustered sensor networks is much more threatening than that of normalsensors. Recently, rekeying schemes which update the exposed keys using the keys unknown to the compromised nodes are emerging. However, they cause some security and efficiency problems such as single group key employment in a cluster, passive eviction of compromised nodes, and excessive communication and computation overhead. In this paper, we present a proactive rekeying scheme using renewals of duster organization for clustered sensor networks. In the proposed scheme, each sensor establishes individual keys with neighbors at network boot-up time, and these keys are employed for later transmissions between sensors and their CH. By the periodic cluster reorganization, the compromised nodes are expelled from network and the individual keys employed in a cluster are changed continuously. Besides, newly elected CHs securely agree a key with sink by informing their members to sink, without exchangingany keying materials. The simulation results shows that the proposed scheme remarkably improves the confidentiality and integrity of data in spite of the increase of compromised nodes. Also, they show that the proposed scheme exploits the precious energy resource more efficiently than SHELL.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2009.05a
• /
• pp.733-736
• /
• 2009

Power Line Communication(PLC) is need not additional communication line. So establishment expense is inexpensive and application is simple. Therefore, lower part network of various application field is possible. However, there are high subordinate interference and noise problem on limited transmission data and communication interference element. Wireless Sensor Network(WSN) is need not infrastructure, Self-regulating network architecture of sensor nodes is possible. So at short time, network construction is available. But, power consumption is increased by active sensing for QoS elevation and unnecessary information transmission, low electric power design and necessity of improve protocol are refered to life shortening problem and is studied. In this paper, supplement problem of power line communication and wireless sensor network mutually and because advantage becomes linkage integration control system using synergy effect of two technologies as more restriction be and tries to approach structurally control network that is improved for smooth network environment construction. Honeywell's hybrid sensor network does comparative analysis(benchmarking). Confirm performance elevation proposing teaming of power line communication and wireless sensor network. Through simulation, service delay decreases and confirms that performance elevation.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.4
• /
• pp.29-35
• /
• 2020

In this paper, we propose a new structured entity recognition DeNERT model. Recently, the field of natural language processing has been actively researched using pre-trained language representation models with a large amount of corpus. In particular, the named entity recognition, which is one of the fields of natural language processing, uses a supervised learning method, which requires a large amount of training dataset and computation. Reinforcement learning is a method that learns through trial and error experience without initial data and is closer to the process of human learning than other machine learning methodologies and is not much applied to the field of natural language processing yet. It is often used in simulation environments such as Atari games and AlphaGo. BERT is a general-purpose language model developed by Google that is pre-trained on large corpus and computational quantities. Recently, it is a language model that shows high performance in the field of natural language processing research and shows high accuracy in many downstream tasks of natural language processing. In this paper, we propose a new named entity recognition DeNERT model using two deep learning models, DQN and BERT. The proposed model is trained by creating a learning environment of reinforcement learning model based on language expression which is the advantage of the general language model. The DeNERT model trained in this way is a faster inference time and higher performance model with a small amount of training dataset. Also, we validate the performance of our model's named entity recognition performance through experiments.

• Korean Journal of Soil Science and Fertilizer
• /
• v.34 no.1
• /
• pp.55-63
• /
• 2001

N fertilizer recommendations to optimize with consideration to maximum crop yields, maximum profits, and minimum N losses to ground or runoff water, an advisory system. Nitrogen Expert System (NES), was developed. The system was to estimate the optimal rate of N fertilizer application cornfields in Illinois. NES was constructed using Smart Elements, a knowledge-based system that manages the expertise of human experts. NES was reinforced by addition of the effect of a productivity index (PI), soil organic matter content (SOM), and pre-sidedressing of nitrate concentration (PSNT) to the optimal N fertilizer recommendation. NES contains 49 rules, 1 class, 14 objects, and 2 properties. NES was successfully operated, showing N recommendations with inputs of three soil properties including PI, SOM, and PSNT. NES can reduce N loss to the environment, but adherence to the recommendations may also reduce farmers income. Therefore, NES will be more effective by evaluating both environmental damage assessment and other economic agricultural management parameters and other soil physico-chemical parameters.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.24 no.6
• /
• pp.525-538
• /
• 2022

In Korea, safety accidents are continuously occurring due to the aging of underground facilities and lack of systematic management. Moreover, although the underground space is continuously being developed, the current status information is not clearly recorded and managed, so there is a limit to the systematic management of underground facilities. Therefore, this study developed an augmented reality-based system that can effectively maintain and manage underground facilities that are difficult to manage because they are located underground. In order to develop an augmented reality-based underground facility management system, three essential requirements, 'precise localization', 'use of BIM information', and 'ensure usability' were derived and reflected in the system. By utilizing Broadcast-RTK, the positional precision was secured to cm level, and the configuration and attribute information of the BIM was converted into the IFC format to construct a system that could be implemented in augmented reality. It developed an application that can optimize usability. Finally, through simulation, the configuration and attribute information of structures and mechanical systems constituting underground facilities were implemented in augmented reality. In addition, it was confirmed that the accurate and highly consistent augmented reality system works even in harsh environment (near high-rise building).

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.31 no.5
• /
• pp.1001-1010
• /
• 2021

Smart factories are complex systems which combine latest information technology (IT) with operation technology (OT). A smart factory aims to provide manufacturing capacity improvement, customized production, and resource reduction with these complex technologies. Although the smart factory is able to increase the efficiency through the technologies, the security level of the whole factory is low due to the vulnerability transfer from IT. In addition, the response and restoration of the business continuity plan are insufficient in case of damage due to the absence of factory security experts. The cope with the such problems, we propose an information security practice content for analyzing the damage by generating ransomware attacks in a digital twin-based smart factory similar to the real world. In our information security content, we introduce our conversion technique of physical devices into virtual machines or simulation models to build a practical environment for the digital twin. This content generates two types of the ransomware attacks according to a defined scenario in the digital twin. When the two generated attacks are successfully completed, at least 8 and 5 of the 23 virtual elements are take damage, respectively. Thus, our proposed content directly identifies the damage caused by the generation of two types of ransomware in the virtual world' smart factory.