• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.30 no.4D
• /
• pp.433-442
• /
• 2010

Recent decrease in the availability of experienced skilled labor and a corresponding lack of new entrants has required the need for automating many of the construction equipment used in the construction industry. In particular, excavators are widely used throughout earthwork operations and automating its tasks enables work to be performed with higher productivity and safety. This paper introduces an optimal path generation method which is one of the core technologies required to make "Intelligent" excavators a reality. The method divides a given earthwork area into unit cells, identifies networks created by linking these cells, and identifies the optimal path an excavator should follow to minimize its total transportation costs. In addition, the method also accounts for drainage direction and path continuity to ensure that the generated path considers site specific conditions.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2022.05a
• /
• pp.321-321
• /
• 2022

유역 모델링은 유역에 강우의 유출 과정을 재현할 수 있는 과정이다. 과거 유역 모델링은 1차원 수준에서 유출과정을 모의하는데 그쳤으나, 기술이 발전함에 있어 입력하는 매개변수 수가 증가함에 따라 모의값의 신뢰성이 높아지고 있다. 본 연구에서는 다양한 매개변수 중 공간매개변수로써 널리 활용되고 있는 수치표고모델의 신뢰성과 범용성을 확인하고자 한다. 유역모델링 연구에 있어, 수치표고모델 정확성은 결과값의 신뢰성을 좌우하는 중요한 인자 중 하나이다. 수치표고모델이란 실제 지형이 나타내는 표고 정보를 수치화 하여 격자 안에 담은 형태의 파일로 대표적으로 DEM(Digital Elevation Models)과 DSM(Digital Surface Models)로 나눌 수 있다 DEM의 경우 해당 지형의 고도정보만을 담고있으며, DSM은 지표면 상의 나무, 건물 등 포함한 지표면의 고도를 담고 있다. 현재 NASA에서는 전 지구의 30m격자 크기로 SRTM-DSM을 제공하고 있으며, 우리나라 국토지리정보원에서도 90m 격자크기의 DEM을 제공하고 있다. 본 연구에서는 남한강 유역의 수치표고모델을 세 가지 Case로 나눠서 유출량 변화 검토를 진행하였다. 산지가 많은 남한강 유역의 10개의 소유역을 선정하였고, 다음과 같이 3개의 Case를 적용하였다. Case1, DEM 자료를 입력했을 경우, Case2, DSM 자료를 입력했을 경우, Case3 DSM+DEM 자료를 입력했을 경우, 각 Case에 대해 유출량을 산정하였고, 그 결과값을 분석하였다. 해당 유역에 세 가지 Case 모두 유출량의 변화량의 큰 차이를 보이지 않았으며, 공간매개변수 적용에 있어 타당성을 보였다. 따라서 본 연구는 인공위성을 통해 산출된 수치표고모델의 신뢰성을 확인 하였고, 활용가능성을 검토 하였다. 이에 따라 향후 연구에 수치표고모델 적용에 있어 미계측유역에도 활용가능한 연구로 기여할 수 있을 것으로 판단된다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2023.07a
• /
• pp.541-544
• /
• 2023

LLM(Largescale Language Model)의 성능 향상을 위한 비용 효율적인 방법으로 ChatGPT, GPT-4와 같은 초거대 모델의 output에 대해 SLM(Small Language Model)을 finetune하는 방법이 주목받고 있다. 그러나, 이러한 접근법은 주로 범용적인 지시사항 모델을 위한 학습 방법으로 사용되며, 제한된 특정 도메인에서는 추가적인 성능 개선의 여지가 있다. 본 연구는 특정 도메인(Writing Assistant)에서의 성능 향상을 위한 새로운 방법인 Self-Guided Approach를 제안한다. Self-Guided Approach는 (1) LLM을 활용해 시드 데이터에 대해 도메인 특화된 metric(유용성, 관련성, 정확성, 세부사항의 수준별) 점수를 매기고, (2) 점수가 매겨진 데이터와 점수가 매겨지지 않은 데이터를 모두 활용하여 supervised 방식으로 SLM을 미세 조정한다. Vicuna에서 제안된 평가 방법인, GPT-4를 활용한 자동평가 프레임워크를 사용하여 Self-Guided Approach로 학습된 SLM의 성능을 평가하였다. 평가 결과 Self-Guided Approach가 Self-instruct, alpaca와 같이, 생성된 instruction 데이터에 튜닝하는 기존의 훈련 방법에 비해 성능이 향상됨을 확인했다. 다양한 스케일의 한국어 오픈 소스 LLM(Polyglot1.3B, PolyGlot3.8B, PolyGlot5.8B)에 대해서 Self-Guided Approach를 활용한 성능 개선을 확인했다. 평가는 GPT-4를 활용한 자동 평가를 진행했으며, Korean Novel Generation 도메인의 경우, 테스트 셋에서 4.547점에서 6.286점의 성능 향상이 발생했으며, Korean scenario Genration 도메인의 경우, 테스트 셋에서 4.038점에서 5.795 점의 성능 향상이 발생했으며, 다른 유사 도메인들에서도 비슷한 점수 향상을 확인했다. Self-Guided Approach의 활용을 통해 특정 도메인(Writing Assistant)에서의 SLM의 성능 개선 가능성을 확인했으며 이는 LLM에 비용부담을 크게 줄이면서도 제한된 도메인에서 성능을 유지하며, LLM을 활용한 응용 서비스에 있어 실질적인 도움을 제공할 수 있을 것으로 기대된다.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.36 no.5
• /
• pp.347-354
• /
• 2023

In this study, virtual reality software was developed to support the generation of an analysis model of a Hanok and to increase the use of the completed analysis model. The structural analysis model of the Hanok was generated using midas Gen, a general-purpose structural analysis software. After converting it into a text-based input file, the developed software stores the data necessary for the examination of the analysis model. Then, in the developed virtual reality software, the three-dimensional analysis model of the Hanok can be visualized in various ways and the related data can be shown by selecting a specific member. Through this process, errors in the analysis model can be identified and corrected to build a complete analysis model. The developed software was applied to three Hanok cases to verify its applicability and effectiveness. The software is expected to be used in other fields besides the structural field.

• Journal of Advanced Navigation Technology
• /
• v.26 no.6
• /
• pp.441-447
• /
• 2022

In this paper, we propose a FLL-assisted-PLL based time synchronization algorithm for telemetry systems where frequency and phase errors exist in time synchronization pulse. The telemetry system may analyze the flight state by acquiring the state information in the distributed system. Therefor, in order to collect each state information without errors, precise time synchronization between the master and the slave is required. At this time, the master's time pulse have frequency and phase changes that can be caused by external and internal factors, so a method to maintain precision time synchronization is essential to provide telemetry data continuously. We propose the FLL-assisted-PLL based algorithm that is capable of high-speed synchronization and has high time synchronization accuracy. The proposed algorithm is verified through python simulation, and the VHDL Logic has been implemented in FPGA to check the performance according to the frequency errors and phase errors.

• Journal of the Society of Disaster Information
• /
• v.19 no.4
• /
• pp.747-756
• /
• 2023

Purpose: In this study, a grout material mixed using non-alkaline silica-based materials, which is an eco-friendly injection material to stabilize ground, is investigated to improve conventional problems. Method: The homogel specimens of Eco-Friendly Non-Alkaline Silica Sol (ENASS) and L.W. and S.G.R., representative silicate grouting are manufactured. Physicochemical and engineering properties of the specimens are evaluated in laboratory with uniaxial compression strength, hydraulic conductivity, shrinkage, chemical resistance, elution, fish poison, waste leaching. Result: Laboratory test results show that the ENASS was superior in all aspects compared to the existing injection matirial. The suitability of the grout material with ENASS is investigated with filed tests. Conclusion: The results of laboratory and field tests demonstrates that the grout material with ENASS is eco-friendly material that increases the strength, decreases the permeability, and discharges pollutants without leaching.

• Journal of the Microelectronics and Packaging Society
• /
• v.30 no.4
• /
• pp.69-78
• /
• 2023

With the rapid growth of artificial intelligence, the demand for semiconductors is enormously increasing everywhere. To ensure the manufacturing quality and quantity simultaneously, the importance of automatic defect detection during the packaging process has been re-visited by adapting various deep learning-based methodologies into automatic packaging defect inspection. Deep learning (DL) models require a large amount of data for training, but due to the nature of the semiconductor industry where security is important, sharing and labeling of relevant data is challenging, making it difficult for model training. In this study, we propose a new framework for securing sufficient data for DL models with fewer computing resources through a divide-and-conquer approach. The proposed method divides high-resolution images into pre-defined sub-regions and assigns conditional labels to each region, then trains individual sub-regions and boundaries with boundary loss inducing the globally coherent and seamless images. Afterwards, full-size image is reconstructed by combining divided sub-regions. The experimental results show that the images obtained through this research have high efficiency, consistency, quality, and generality.

• Smart Media Journal
• /
• v.12 no.9
• /
• pp.60-71
• /
• 2023

A microservice architecture that accommodates the heterogeneity of various development environments and enables flexible maintenance can secure business agility to manage services in line with rapidly changing requirements. Due to the nature of MSA, where communication between microservices within a service is frequent, the boundary security that has been used in the past is not sufficient in terms of security, and a Zerotrust system is required. In addition, as the size of microservices increases, definition of access control policies according to the API format of each service is required, and difficulties in policy management increase, such as unnecessary governance overhead in the process of redistributing services. In this paper, we propose a microservice architecture that centrally manages policies by separating access control decision and enforcement with a general-purpose policy engine called OPA (Open Policy Agent) for collective and flexible policy management in Zerotrust security-applied environments.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.6
• /
• pp.911-917
• /
• 2023

In today's world, we find ourselves facing energy crises due to factors such as war and climate crises. To prepare for these energy crises, many researchers continue to study systems related to energy monitoring and conservation, such as energy management systems, energy monitoring, and energy conservation. In line with these efforts, nations are making it mandatory for energy-consuming facilities to implement these systems. However, these facilities, limited by space and energy constraints, are exploring ways to improve. This research explores the operation of a data collection system using low-performance embedded devices. In this context, it proves that an optimized version of RocksDB, a Key-Value store, outperforms traditional databases when it comes to time-series data. Furthermore, a comprehensive database evaluation tool was employed to assess various databases, including optimized RocksDB and regular RocksDB. In addition, heterogeneous databases and evaluations are conducted using a UD Benchmark tool to evaluate them. As a result, we were able to see that on devices with low performance, the time required was up to 11 times shorter than that of other databases.

• Convergence Security Journal
• /
• v.24 no.1
• /
• pp.11-20
• /
• 2024

Recently, smart buildings that can efficiently manage energy using ICT technology and operate and control through the building automation system by collecting data from a large number of IoT sensors in real time are attracting attention. However, as data management is carried out through an open environment, the safety of smart buildings is threatened by the security vulnerability of the existing building automation protocol. Therefore, in this paper, we analyze the major data link technology of BACnet, which is used universally, and propose OWE-based efficient BACnet/SC that can ensure the reliability of data in a wireless environment. The proposed protocol enables safe communication even in an open network by applying OWE and provides the same level of security as BACnet/SC in a TLS environment. As a result, it reduces the connection process twice and reduces the average time required by 40%, enabling more efficient communication than before.