• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.961-962
• /
• 2023

포트폴리오 구성문제는 과거부터 현재까지 많은 연구가 이루어지고 있다. 현재는 강화학습을 통해 포트폴리오를 구성하는 연구가 많이 진행되고있다. 포트폴리오를 구성함에 있어 종목선택과 각 종목을 얼만큼 투자할 것인지는 둘 다 중요한 문제이다. 본 연구에서는 과거부터 많이 사용해오던 방식을 차용하여 강화학습 방법과 접목시켰고 이를 통해 설명력이 높은 모델을 만들려고 노력하였다. 강화학습에 사용한 모델은 PPO(Proximal Policy Optimization)을 기본으로 하였고 인공신경망은 LSTM을 활용하였다. 실험결과 실험 기간 동안(2023년 3월 30일 부터 108 영업일 까지)의 코스피 수익률은 5%인데 반해 본 연구에서 제시한 모델의 수익률은 평균 약 9%를 기록했다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.376-377
• /
• 2023

스마트홈은 기술 시스템, 자동화 프로세스, 원격 제어 기기 등을 아파트나 주택에서 사용하는 것을 말한다. 주요 목적은 가정에서 삶의 질과 편의성을 높이는 것이다. 현재의 스마트홈은 사용자의 원격 제어 방식을 사용하고 있다. 이러한 방식은 고정된 시간에만 스마트홈이 작동하도록 한다는 문제가 있었다. 연관분석과 순차분석을 통해 AI가 상황과 사용자의 취향을 학습한다면, 스스로 최적화된 패턴을 제공할 수 있을 것이다.

• Review of KIISC
• /
• v.33 no.6
• /
• pp.65-77
• /
• 2023

산업용 사물 인터넷(IIoT)은 자원 관리 및 최적화, 신속성, 지속가능한 생산, 자동화 등의 특징으로 인해 다양한 산업분 야에서 활발하게 사용되고 있다. 수많은 IIoT 기기에서 발생된 데이터를 처리하는 것은 기존 중앙 처리 시스템에 큰 부담을 주게 된다. 이러한 데이터들의 효과적인 관리를 위해 데이터가 발생한 엣지 기기, 엣지 서버 등 로컬위치에서 실시간으로 프로세스를 실행하여 네트워크 대역폭 절약, 낮은 지연 시간 등 특징을 가진 엣지 컴퓨팅 기술을 사용한다. 하지만, 엣지 컴퓨팅 적용 시, 인터넷과 연결된 IIoT 기기 수 증가, 취약한 IIoT 기기, 분산된 환경으로 인해 공격 표면 확장되어 엣지 컴퓨팅 환경에서의 새로운 보안 위협이 발생할 수 있다. 이에 본 논문에서는 IIoT 및 엣지 컴퓨팅 정의, 아키텍처, 각 산업분야별 적용한 사례에 대해 살펴보고, 엣지 컴퓨팅에서 발생 가능한 보안 위협을 분석하였다. 또한, 엣지 컴퓨팅 기반 IIoT에 대한 각 산업 분야별 보안 연구 동향에 대해서 분석하였다.

• Journal of Navigation and Port Research
• /
• v.39 no.4
• /
• pp.353-360
• /
• 2015

The offshore oil production requires a huge amount of cost and time accompanied by multiple variables due to the peculiar nature of 'offshore'. And every process concerned is controlled by elaborate series of plans for reducing loss of lives, environment and property. This paper treats an optimization problem for offshore oil production and transportation. We present an offshore production and transportation network to define scope of the problem and construct a mixed integer linear programming model to tackle it. To demonstrate the validity of the optimization model presented, some computational experiments based on hypothetical offshore oil fields and demand markets are carried out by using MS Office Excel solver. The downstream of the offshore production and transportation network ends up with the maritime transportation problem distributing the crude oil produced from offshore fields to demand markets. We used MoDiSS(Model-based DSS in Ship Scheduling) which was built to resolve this maritime transportation problem. The paper concludes with the remark that the results of the study might be meaningfully applicable to the real world problems of offshore oil production and transportation.

• Korean Journal of Construction Engineering and Management
• /
• v.22 no.1
• /
• pp.36-46
• /
• 2021

With an increasing trend toward high-rise modular construction, the simultaneous use of tower cranes at a modular construction site has recently been observed. Tower crane layout planning (TCLP) has a significant effect on cost, duration, safety and productivity of a project. In a modular construction project, particularly, poor decision about the layout of tower cranes is likely to have negative effects like additional employment of cranes and redesign, which will lead to additional costs and possible delays. It is, therefore, crucial to conduct thorough inspection of field conditions, lifting materials, tower crane capacity to make decisions on the layout of tower cranes. However, several challenges exist in planning for a multi-crane construction site in terms of safety and collaboration, which makes planning with experience and intuition complicated. This paper suggests a multi-objective optimization model for selection of the number of tower cranes, their models and locations, which minimizes cost and conflict. The proposed model contributes to the body of knowledge by showing the feasibility of using multi-objective optimization for TCLP decision-making process with consideration of trade-offs between cost and conflict.

• Korean Journal of Environmental Biology
• /
• v.35 no.2
• /
• pp.207-214
• /
• 2017

There is little information about the seed longevity of wild plants, although seed bank storage is an important tool for biodiversity conservation. This study was conducted to predict the seed viability equation of Agastache rugosa. The A. rugosa seeds were stored at moisture contents ranging from 2.7 to 12.5%, and temperatures between 10 and $50^{\circ}C$. Viability data were fitted to the seed viability equation in a one step and two step approach. The A. rugosa seeds showed orthodox seed storage behaviour. The viability constants were $K_E=6.9297$, $C_W=4.2551$ $C_H=0.0329$, and $C_Q=0.00048$. The P85 of A. rugosa seeds was predicted to 152 years under standard seed bank conditions. The P85 predicted by seed viability equation can be used as basic information for optimization of seed storage processes.

• Korean Journal of Construction Engineering and Management
• /
• v.16 no.5
• /
• pp.54-64
• /
• 2015

The modular construction has several advantages such as high quality of product, safe work condition and short construction duration. The manufacturing planning of modular construction should consider time frame of manufacturing, transport and erection process with limited resources (e.g., modular units, transporter and workers). The manufacturing planning of multi modular construction project manages the modular construction's characteristics and diversity of projects, as a type of modular unit, modular unit quantities, and date for delivery. However, current modular manufacturing planning techniques are weak in dealing with resource interactions and each project requirement in multi modular construction project environments. Inefficient allocation of resources during multi modular construction project may cause delays and cost overruns to construction operation. In this circumstance, this research suggest a manufacturing planning model for schedule optimization of multi project of modular construction, using genetic algorithm as one of the powerful method for schedule optimization with multiple constrained resources. Comparing to the result of the existed schedule of case study, setting optimized scheduling for multi project decrease the total factory producing schedule. By using proposed optimization tool, efficient allocation of resource and saving project time is expected.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.3
• /
• pp.891-899
• /
• 2024

The Fourth Industrial Revolution is centered on a personalized demand fulfillment economy and is all about transformation and flexible processing that can deliver what customers want in real time across space and time. This paper implements the construction and operation of a packaging platform that can instantly procure the required packaging products based on real-time orders and evaluates its performance. The components of customer satisfaction are flexible and dependent on the situation which requires efficient management of enterprise operational processes based on an e-SCM platform. An OMS optimized for these conditions plays an important role in maximizing and differentiating the efficiency of a company's operations and improving its cost advantage. OMS is a system of mass customization that provides efficient MOT(Moment of Truth) logistics services to meet the eco-friendly issues of many individual customers and achieve optimized logistics operation goals to enhance repurchase intentions and sustainable business. OMS precisely analyzes the collected data to support information and decision-making related to efficiency, productivity, cost and provide accurate reports. It uses data visualization tools to express data visually and suggests directions for improvement of the operational process through statistics and prediction analysis.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.9
• /
• pp.1854-1864
• /
• 2011

Production planning is one of the most important activities in shipbuilding enterprises. Shop-floor supervisors and planners still do not have enough information to effectively analyze shop operations because of the difference between production planning and shop-floor scheduling. In this paper, process analysis was conducted between production planning and shop-floor control to clarify the difference, and the necessity of the manufacturing execution system(MES) was derived in a shipyard. Therefore, the simulation-based ship production execution system(SPEXS) was defined by analyzing characteristics of MES. The architectural functions of the system were deducted from the process of requirement analysis. The SPEXS' framework was constructed on the basis of the architectural functions. This framework will provide more reliable production schedules and allow engineers to plan and control shop operations in real-time.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.37 no.7
• /
• pp.669-674
• /
• 2013

A three-dimensional hydrogel scaffold has been proposed for the effective production of biomimetic intestinal villi to reduce ethical and cost problems caused by animal testing in pharmaceutical development. This study explores an experimental approach to develop a new technique based on laser machining and microdrilling processes to produce a plastic mold for the fabrication of a three-dimensional hydrogel scaffold. For process optimization, both the laser machining and the microdrilling experiments are conducted by varying the experimental conditions, and structural characterization of the mold and intestinal villi were performed using SEM (scanning electron microscope) and OM (optical microscope) images. Polycarbonate (PC) was used as a candidate material. The experimental results show that intestinal villi can be fabricated by both laser and microdrilling machining processes.