• 국제학술발표논문집
• /
• The 3th International Conference on Construction Engineering and Project Management
• /
• pp.409-416
• /
• 2009

This paper presents a tool called Quality-Cost optimization system (QCOS), which integrates Multi-Objective Genetic Algorithm (MOGA) and Quality Function Deployment (QFD), for tradeoff between quality and cost of the unitized metal curtain-wall unit. A construction owner as the external customer pursues to maximize the quality of the curtain-wall unit. However, the contractor as the internal customer pursues to minimize the cost involved in designing, manufacturing and installing the curtain-wall unit. It is crucial for project manager to find the tradeoff point which satisfies the conflicting interests pursued by the both parties. The system would be beneficial to establish a quality plan satisfying the both parties. Survey questionnaires were administered to the construction owner who has an experience of curtain-wall project, the architects who are the independent assessor, and the contractors who were involved in curtain-wall design and installation. The Customer Requirements (CRs) and their importance weights, the relationship between CRs and Technical Attributes (TAs) consisting of a curtain-wall unit, and the cost ratios of each components consisting curtain-wall unit are obtained from the three groups mentioned previously. The data obtained from the surveys were used as the QFD input to compute the Owner Satisfaction (OS) and Contractor Satisfaction (CS). MOGA is applied to optimize resource allocation under limited budget when multi-objectives, OS and CS, are pursued at the same time. The deterministic multi-objective optimization method using MOGA and QFD is extended to stochastic model to better deal with the uncertainties of QFD input and the variability of QFD output. A case study demonstrates the system and verifies the system conformance.

• 한국강구조학회 논문집
• /
• 제17권6호통권79호
• /
• pp.661-672
• /
• 2005

본 연구에서는 많은 양의 함수 계산을 요구하는 확률론적 최적화 기법을 보다 효과적으로 강구조물에 적용하여 수행하고자 한다. 다양한 과학, 응용공학 분야에서 많은 시간이 소요되는 과정을 대체하는데 효과적인 도구로 출현한 인공신경망을 최적화 과정 중 많은 수의 유한요소 해석이 요구되는 재해석 문제에 적용함으로서 유한요소법의 평형방정식의 해의 근사해를 추정하여 재해석과정을 보다 간단하고 용이하게 수행하고자 한다. 또한 이용된 인공신경망의 학습효과의 개선을 위해 유전알고리즘을 적용한다. 확률론적 구조최적화 기법으로는 진화론적 방법에 기초한 알고리즘을 사용한다. 수치 예로써 전형적인 체적(중량)문제와 실 경비함수를 목적함수로 갖는 강구조물 모형에 본 연구의 알고리즘을 적용하여 본 알고리즘의 적용성과 타당성을 증명하였다.

• Geomechanics and Engineering
• /
• 제31권6호
• /
• pp.545-556
• /
• 2022

Evaluation and optimization of tunnel wall convergence (TWC) plays a vital role in preventing potential problems during tunnel construction and utilization stage. When convergence occurs at a high rate, it can lead to significant problems such as reducing the advance rate and safety, which in turn increases operating costs. In order to design an effective solution, it is important to accurately predict the degree of TWC; this can reduce the level of concern and have a positive effect on the design. With the development of soft computing methods, the use of deep learning algorithms and neural networks in tunnel construction has expanded in recent years. The current study aims to employ the long-short-term memory (LSTM) deep neural network predictor model to predict the TWC, based on 550 data points of observed parameters developed by collecting required data from different tunnelling projects. Among the data collected during the pre-construction and construction phases of the project, 80% is randomly used to train the model and the rest is used to test the model. Several loss functions including root mean square error (RMSE) and coefficient of determination (R²) were used to assess the performance and precision of the applied method. The results of the proposed models indicate an acceptable and reliable accuracy. In fact, the results show that the predicted values are in good agreement with the observed actual data. The proposed model can be considered for use in similar ground and tunneling conditions. It is important to note that this work has the potential to reduce the tunneling uncertainties significantly and make deep learning a valuable tool for planning tunnels.

• 대한토목학회논문집
• /
• 제35권1호
• /
• pp.165-172
• /
• 2015

철도 노선설계 후 노선의 적합도 평가는 수요의 변화를 기준으로 한 장래 가치의 환산 또는 비용편익 분석을 통한 대안 간 평가 등으로 이루어지고 있다. 그러나 이는 설계 가능한 모든 대안의 조사가 불가능하고, 설계기준과 같은 기본 틀은 편차가 적어 대안간 적합성을 상호비교하는 데에는 적절치 못할 것이다. 또한 비용은 설계요소에 따라 변화되므로 이를 적용하는 데에는 한계가 있다. 본 연구에서는 자동화된 설계방법으로 조사된 모든 대안을 대상으로 하고, 설계기준을 모든 대안에 적용하여 기준을 벗어나지 않도록 하며, 단일 평가항목이 아닌 다 기준 의사결정 방법을 사용하여 선형설계요소의 가중치를 평가하였다. AHP (Analytic Hierarchical Process) 기법을 사용한 가중치 평가결과를 통하여 철도선형계획에서 중점을 두어야 하는 요소들을 고려함으로써, 보다 최적화된 철도선형을 도출할 수 있을 것으로 기대된다.

• 비파괴검사학회지
• /
• 제26권6호
• /
• pp.403-410
• /
• 2006

Eddy current testing (ECT) signals produced by a differential bobbin coil probe vary according to probe design parameters such as the number of turns, geometry and coil gap size. In the present study, the characteristics of a differential bobbin coil probe signals are investigated by numerical simulation in order to determine the optimum coil gap. For verification of numerical simulation accuracy, a specially designed bobbin probe of which the coil gap can be adjusted is fabricated and a series of experiments to acquire signals from two kinds of standard tubes with the variation in coil gap is performed. Then, the experimental signals are compared to the simulation results. Based on this investigation, a decision on the optimum range of coil gap is made. The theoretically predicted signals agree very well to the experimental signals. In fact, this excellent agreement demonstrates a high potential of the simulation as a design optimization tool for ECT bobbin probes.

• 산경연구논집
• /
• 제8권6호
• /
• pp.7-16
• /
• 2017

Purpose - Today's world revolves around search engines which are the driving force behind any marketer. The thirst for marketing has led to the evolution of online 'Pay per click' over last few years and is the most widely used instrument. Research design, data, and methodology - Exploratory research design highlights many marketing variables getting affected by pay per click marketing. To analyze the said phenomenon, the data was gathered through questionnaire from the sample of 338 respondents which were selected by simple random sampling method mostly from the National Capital Region (NCR) of Delhi in India. The data collected from the respondents was loaded on SAS base for exploratory factor analysis and multiple regression analysis. Results - Pay per click as a marketing tool has significant impact on the consumers. The most prominent factors of pay per click marketing identified in the research are Ad quality, Competition, Targeting, Trend and Budget. Conclusions - Organic as well as inorganic ads, keeping in mind the end goal to gage the exchange of these two postings in the marked look territory. Additionally, here we dissected supported pursuit promotions in all. It would be beneficial to break down the impact of promotion position on the pay per click marketing.

• 제어로봇시스템학회논문지
• /
• 제19권4호
• /
• pp.299-305
• /
• 2013

The complexity and severe nonlinearity of multi-machine power systems make it difficult to design a control input for voltage regulation using modern control theory. This paper presents a model-based PID control scheme for the regulation of the bus voltage to a desired value. To this end, a fourth-order linear system is constructed using input and output data obtained using the TSAT (Transient Security Assessment Tool); the input is assumed to be applied to the grid through the STATCOM (STATic synchronous COMpensator) and the output from the grid is a bus voltage. On the basis of the model, it is identified as to which open-loop poles of the system make the response to a step input oscillatory. To reduce this oscillatory response effectively, a model-based PID control is designed in such a way that the oscillatory poles are no longer problematic in the closed loop. Simulation results show that the proposed PID control dampens the response effectively.

• 대한기계학회논문집B
• /
• 제28권1호
• /
• pp.16-23
• /
• 2004

Acoustic characteristics of unbaffled and baffled combustion chamber are experimentally investigated under atmospheric condition to preliminarily determine baffle for mitigation of combustion instability. To investigate the effect of the baffle which has several configurations such as radial baffles and hub/blade baffle, resonant-frequency shift and damping factors of the chamber were analyzed and compared quantitatively with those of the unbaffled combustion chamber. From a view of acoustic characteristics, radial baffles with several configurations have not much difference in resonant-frequency shift and damping factor ratio with each other. On the other hand, hub and blade baffle is very effective to suppress the first tangential mode which was found to be the most harmful acoustic mode in KSR(Korean Sounding Rocket)-III engine. But more study on design parameters such as hub size and axial length should be done for complete optimization of hub and blade baffle. The present study based on linear acoustic analysis is expected to be a useful confirming tool to predict acoustic field and design a passive control devices such as baffle and acoustic cavity.

• 대한기계학회논문집A
• /
• 제24권10호
• /
• pp.2543-2551
• /
• 2000

Precision injection molding is an important technology for improving productivity and lowering costs in the fields of medical components, lenses and electrical connectors. The quality of injection molded parts is affected by various processing conditions such as filling time and packing pressure profile. It is difficult to consider all the variables at the same time for prediction of the quality. In this study, the genetic algorithm was used to obtain the optimal processing conditions for minimizing the volumetric shrinkage of molded parts. For a higher convergence rate, the method of design of experiments was used to analyze the relationship between processing conditions and volumetric shrinkage of molded parts, which served as analysis tool for the capability of searching optimal processing conditions but also greatly reduces the calculation time by utilizing the information of searching area. As a practical example, compact disks that require micron-level precision were chosen for the study.

• International Journal of Precision Engineering and Manufacturing
• /
• 제2권4호
• /
• pp.5-11
• /
• 2001

Vibration and noise of gears is doc to the transmission error and the vibration exciting force caused by the periodically alternating tooth stiffness. Transmission error is the rotation delay between driving and driven gear caused by manufacturing error, alignment error in assembly and so on. Tooth stiffness changes with the proceeding mesh of teeth. The purpose of this study is to develop how to calculate simultaneously the optimum amounts of tooth profile modification. end relief and crowning by minimizing the vibration exciting force of helical gears. We estimate the vibration exciting force by the meshing analysis of gears. Formulated constraints of this problem consist of contact ratio and strengths of gear teeth such as tooth bending strength, surface durability, and scoring. ADS(Automated Design Synthesis) is used as an optimization tool. We also investigate the relation between the aspect ratio and the optimum values of tooth modification. The proposed method can calculate the optimum amount of tooth modification automatically and is expected to be practically useful to resolve the problem of vibration of helical gears.