• Nuclear Engineering and Technology
• /
• 제55권12호
• /
• pp.4607-4616
• /
• 2023

We implement machine learning regression models to predict peak pressures of primary and secondary systems, a major safety concern in Loss Of Condenser Vacuum (LOCV) accident. We selected the Multi-dimensional Analysis of Reactor Safety-KINS standard (MARS-KS) code to analyze the LOCV accident, and the reference plant is the Korean Optimized Power Reactor 1000MWe (OPR1000). eXtreme Gradient Boosting (XGBoost) is selected as a machine learning tool. The MARS-KS code is used to generate LOCV accident data and the data is applied to train the machine learning model. Hyperparameter optimization is performed using a simulated annealing. The randomly generated combination of initial conditions within the operating range is put into the input of the XGBoost model to predict the peak pressure. These initial conditions that cause peak pressure with MARS-KS generate the results. After such a process, the error between the predicted value and the code output is calculated. Uncertainty about the machine learning model is also calculated to verify the model accuracy. The machine learning model presented in this paper successfully identifies a combination of initial conditions that produce a more conservative peak pressure than the values calculated with existing methodologies.

• 대한조선학회논문집
• /
• 제53권5호
• /
• pp.428-434
• /
• 2016

A lot of drawings for pipe construction and installation are needed to construct plant process system on the offshore plant structures. Depending on their scale or complexity, the required number of drawings related pipes sometimes amounts to several hundreds of thousands. Most major shipyards, therefore, have their own system which can automatically depict pipes’ geometric, manufacturing, and BOM(Bill of Material) information on the drawings. However, as the complexity and absolute quantity in the isometric region on the drawings is increased, the information extraction from the customized DB and positioning at the typical locations does not get to be enough to avoid the overlap between geometric contours, labels, and symbols. For this reason, the novel methods to arrange additional annotations without overlaps are presented in the paper. This approach is expected to increase the readability and legibility of the drawing and prevent the human error, and finally decrease the time-consuming and tedious jobs which are unnecessarily required to designers.

• 대한산업공학회지
• /
• 제35권4호
• /
• pp.257-265
• /
• 2009

This paper considers the hybrid flow shop scheduling problem for the objective of minimizing the number of tardy jobs. In hybrid flow shops, each job is processed through multiple production stages in series, each of which has multiple identical parallel machines. The problem is to determine the allocation of jobs to the parallel machines at each stage as well as the sequence of the jobs assigned to each machine. Due to the complexity of the problem, we suggest search heuristics, tabu search and simulated annealing algorithms with a new method to generate neighborhood solutions. In particular, to evaluate and select neighborhood solutions, three surrogate objectives are additionally suggested because not much difference in the number of tardy jobs can be found among the neighborhoods. To test the performances of the surrogate objective based search heuristics, computational experiments were performed on a number of test instances and the results show that the surrogate objective based search heuristics were better than the original ones. Also, they gave the optimal solutions for most small-size test instances.

• 비파괴검사학회지
• /
• 제34권2호
• /
• pp.119-127
• /
• 2014

An epoxy coating applied to the concrete surface of a containment building deteriorates in hazardous environments such as those containing radiation, heat, and moisture. Unlike metals, the epoxy coating on a concrete liner absorbs and discharges moisture during the degradations process, so it has a different density and volume during service. In this study, acoustic impedance was adopted for characterizing the degradation of a glass-reinforced epoxy coating using the acoustic reflection coefficient (reflectance) on a rough epoxy coating. For estimating the acoustic reflectance on a wavy epoxy coating surface, a probabilistic model was developed to represent the multiple irregular reflections of the acoustic wave from the wavy surface on the basis of the simulated annealing technique. A number of epoxy-coated concrete specimens were prepared and exposed to accelerated aging conditions to induce an artificial aging degradation in them. The acoustic impedance of the degraded epoxy coating was estimated successfully by minimizing the error between a waveform calculated from the mathematical model and a waveform measured from the surface of the rough coating.

• 한국자기공명학회논문지
• /
• 제19권1호
• /
• pp.1-10
• /
• 2015

Refinements with atomistic molecular dynamics (MD) simulation have contributed to improving the qualities of NMR structures. In most cases, the calculations with atomistic MD simulation for NMR structures employ generalized-Born implicit solvent model (GBIS) to take into accounts solvation effects. Developments in algorithms and computational capacities have ameliorated GBIS to approximate solvation effects that explicit solvents bring about. However, the quantitative comparison of NMR structures in the latest GBIS and explicit solvents is lacking. In this study, we report the direct comparison of NMR structures that atomistic MD simulation coupled with GBIS and water molecules refined. Two model proteins, GB1 and ubiquitin, were recalculated with experimental distance and torsion angle restraints, under a series of simulated annealing time steps. Whereas the root mean square deviations of the resulting structures were apparently similar, AMBER energies, the most favored regions in Ramachandran plot, and MolProbity clash scores witnessed that GBIS-refined structures had the better geometries. The outperformance by GBIS was distinct in the structure calculations with sparse experimental restraints. We show that the superiority stemmed, at least in parts, from the inclusion of all the pairs of non-bonded interactions. The shorter computational times with GBIS than those for explicit solvents makes GBIS a powerful method for improving structural qualities particularly under the conditions that experimental restraints are insufficient. We also propose a method to separate the native-like folds from non-violating diverged structures.

• 한국광학회지
• /
• 제14권4호
• /
• pp.377-382
• /
• 2003

본 논문에서는 이진 위상 홀로그램과 간섭계를 이용하여 디지털 데이터의 저작권을 보호할 수 있는 새로운 워터마킹 방법을 제안하였다. 먼저, SA알고리듬을 이용하여 숨겨질 마크영상의 이진 위상 홀로그램을 설계하였다. 이를 원 영상에 선형중첩 시킴으로써 워터마킹된 영상을 얻었다. 워터마킹된 영상에서 마크영상을 추출하기 위한 복호화 과정은 경로차가 같도록 구성된 마흐-젠더 간섭계를 이용하여 워터마킹된 영상의 위상변조 성분과 위상변조된 원 영상과의 간섭세기론 역푸리에 변환 함으로써 얻었다. 제안된 방법은 임의의 절단된 영상에 대해서 강인함을 컴퓨터 모의실험을 통해 확인하였으며, 위상 변조 특성이 있는 LCD를 사용하여 광학적으로 구현하였다.

• 한국항해항만학회:학술대회논문집
• /
• 한국항해항만학회 2000년도 추계학술대회논문집
• /
• pp.131-142
• /
• 2000

Facility layout is the early stage of system design that requires a mid-term or long-term plan. Since improper facility layout might incur substantial logistics cost including material handling and re-installment costs, due consideration must be given to decisions on facility layout. Facility layout is concerned with low to arrange equipment necessary for production in a given space. Its objective is to minimize the sum of all the products of each equipment's amount of flow multiplied by distance. Facility layout also is related to the issue of NP-complete, i.e., calculated amounts exponentially increase with the increase of the number of equipment. This study discusses Hybrid GA developed, as an algorithm for facility layout, to solve the above-mentioned problems. The algorithm, which is designed to efficiently place equipment, automatically produces a horizontal passageway by the block, if a designer provides the width and length of the space to be handled. In addition, this study demonstrates the validity of the Algorithm by comparing with existing algorithms that have been developed. We present a Hybrid GA approach to the facility layout problem that improves on existing work in terms of solution quality and method. Experimental results show that the proposed algorithm is able to produce better solution quality and more practical layouts than the ones obtained by applying existing algorithms.

• Structural Engineering and Mechanics
• /
• 제60권5호
• /
• pp.795-807
• /
• 2016

This study tries to examine the effect of different parameters on stress analysis of infinite plates with central quasi-triangular cutout using particle swarm optimization (PSO) algorithm and also an attempt has been made to introduce general optimum parameters in order to achieve the minimum amount of stress concentration around this type of cutout on isotropic and orthotropic plates. Basis of the presented method is expansion of analytical method conducted by Lekhnitskii for circular and elliptical cutouts. Design variables in this study include fiber angle, load angle, curvature radius of the corner of the cutout, rotation angle of the cutout and at last material of the plate. Also, diagrams of convergence and duration time of the desired problem are compared with Simulated Annealing algorithm. Conducted comparison is indicative of appropriateness of this method in optimization of the plates. Finite element numerical solution is employed to examine the results of present analytical solution. Overlap of the results of the two methods confirms the validity of the presented solution. Results show that by selecting the aforementioned parameters properly, less amounts of stress can be achieved around the cutout leading to an increase in load-bearing capacity of the structure.

• 한국항해항만학회:학술대회논문집
• /
• 한국항해항만학회 2006년도 International Symposium on GPS/GNSS Vol.2
• /
• pp.329-334
• /
• 2006

Rail clamps are mechanical components installed to fix the container crane to its bottoms from wind blast or slip. Rail clamps should be designed to survive the harsh wind loading condition. In this study, the jaw structure that is one part of wedge-typed rail clamp is optimized, considering strength under the severe wind loading condition. According to the classification of structural optimization, the structural optimization of a jaw belongs to shape optimization. In the conventional structural optimization methods, they have difficulties in defining complex shape design variables and preventing mesh distortions. To overcome the difficulties, the metamodel using kriging interpolation method is introduced, replacing true response by approximate one. This research presents the shape optimization of a jaw using iterative kriging interpolation models and simulated annealing algorithm. The new kriging models are iteratively constructed by refining the former kriging models. This process is continued until the convergence criteria are satisfied. The optimum results obtained by the suggested method are compared with those obtained by the DOE (design of experiments) and VT (variation technology) methods built in ANSYS WORKBENCH.

• 한국기계가공학회지
• /
• 제7권3호
• /
• pp.17-24
• /
• 2008

Rail clamps are mechanical components installed to fix the container crane to its lower members against wind blast or slip. According to rail clamps should be designed to survive harsh wind loading conditions. In this study, a jaw structure, which is a part of a wedge-typed rail clamp, is optimized with respect to its strength under a severe wind loading condition. According to the classification of structural optimization, the structural optimization of a jaw is included in the category of shape optimization. Conventional structural optimization methods have difficulties in defining complex shape design variables and preventing mesh distortions. To overcome the difficulties, the metamodel using Kriging interpolation method is introduced to replace the true response by an approximate one. This research presents the shape optimization of a jaw using iterative Kriging interpolation models and a simulated annealing algorithm. The new Kriging models are iteratively constructed by refining the former Kriging models. This process is continued until the convergence criteria are satisfied. The optimum results obtained by the suggested method are compared with those obtained by the DOE (design of experiments) and VT (variation technology) methods built in ANSYS WORKBENCH.