• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 1993년도 추계학술대회 논문집
• /
• pp.666-671
• /
• 1993

Fatigue life prediction under multi-axial variable load were performed for Aluminium 7075-T651 alloy using SAE Notched specimen & Torque tube shaft component specimen. When variable multiaxial load is applied to material, maximum damaged plane(critical plane) change. To clarify the situation, experiment is performed on two different changing load path. For multiaxial fatigue life prediction, miner rule is expanded to critical plane theory. Shear based parameter and Elliptical parameter give better correlation. This suggests that miner rule can be applicable on multi-axial variable load.

• 환경영향평가
• /
• 제22권6호
• /
• pp.713-724
• /
• 2013

In terms of waste load allocation, inequality of waste load discharge must be considered as well as economic aspects such as minimization of waste load abatement. The inequality of waste load discharge between areas was calculated with Gini coefficient and was included as one of the objective functions of the multi-objective waste load allocation. In the past, multi-objective functions were usually weighted and then transformed into a single objective optimization problem. Recently, however, due to the difficulties of applying weighting factors, multi-objective genetic algorithms (GA) that require only one execution for optimization is being developed. This study analyzes multi-objective waste load allocation using NSGA-II-aJG that applies Pareto-dominance theory and it's adaptation of jumping gene. A sensitivity analysis was conducted for the parameters that have significant influence on the solution of multi-objective GA such as population size, crossover probability, mutation probability, length of chromosome, jumping gene probability. Among the five aforementioned parameters, mutation probability turned out to be the most sensitive parameter towards the objective function of minimization of waste load abatement. Spacing and maximum spread are indexes that show the distribution and range of optimum solution, and these two values were the optimum or near optimal values for the selected parameter values to minimize waste load abatement.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권3호
• /
• pp.911-931
• /
• 2021

Recently, most cloud services use Docker container environment to provide their services. However, there are no researches to evaluate the performance of communication libraries for multi-GPU based distributed deep learning in a Docker container environment. In this paper, we propose an efficient communication architecture for multi-GPU based deep learning in a Docker container environment by evaluating the performances of various communication libraries. We compare the performances of the parameter server architecture and the All-reduce architecture, which are typical distributed deep learning architectures. Further, we analyze the performances of two separate multi-GPU resource allocation policies - allocating a single GPU to each Docker container and allocating multiple GPUs to each Docker container. We also experiment with the scalability of collective communication by increasing the number of GPUs from one to four. Through experiments, we compare OpenMPI and MPICH, which are representative open source MPI libraries, and NCCL, which is NVIDIA's collective communication library for the multi-GPU setting. In the parameter server architecture, we show that using CUDA-aware OpenMPI with multi-GPU per Docker container environment reduces communication latency by up to 75%. Also, we show that using NCCL in All-reduce architecture reduces communication latency by up to 93% compared to other libraries.

• International Journal of Automotive Technology
• /
• 제6권3호
• /
• pp.285-291
• /
• 2005

In this paper, the author proposes a fatigue damage parameter of spot welded joints under proportional loading. The proposed fatigue damage parameter is developed based on von Mises' equivalent stress and local structural stress at the edge of spot weld nugget. The structural stress at the edges of the weld nugget in each sheet is calculated using the forces and moments that are determined by finite element analysis. A structural equivalent stress is then calculated by von Mises' equivalent stress equation. The structural equivalent stresses are correlated to experimental fatigue life of the spot welded joints. The proposed parameter is evaluated with fatigue test data of spot welds subjected to multi axial and tensile-shear loads. Sheppard's parameter and Rupp and co-workers' parameter are also evaluated with the same test data to compare with the author's parameter. This proposed parameter presents a better correlation with experimental fatigue data than those of Sheppard's and Rupp and co-workers' parameter. The proposed parameter should be very effective for durability calculations during the early design phase since coarsely meshed finite element models can be employed.

• 대한전자공학회논문지
• /
• 제26권8호
• /
• pp.1187-1196
• /
• 1989

In this paper, a method for computing the capacitance parameter for a multi-interconnection line in a multilayered dielectric region is presented. The number of interconnection lines and the number of dielectric layers are arbitrary, and the interconnection lines are finite cross section or infinite cross section. The surface of lines and dielectric interface are divided into subsection. The surface charge density of each subsection is a constant step-pulse function for each subsection. After the solution of surface charge density is effected by the method of moments, capacitance parameter is calculated.

• 한국산업융합학회 논문집
• /
• 제7권2호
• /
• pp.215-220
• /
• 2004

In this paper, some algorithms for robust stabilization of linerar time - invariant single - input - multi output (SIMO) systems subject to parameter perturbatations are presented. At first, the determination algorithm of the largest stable hypersphere in the parameter space of a given characteristic polynomial with its coefficient perturbations near some stable nominal values is presented. These algorithms iteratively enlarge the stability hypersph ere in plant parameter space and can be used to design a controller to stabilize a plant subject to givien range of parameter ecxursions.

• Smart Structures and Systems
• /
• 제14권5호
• /
• pp.831-845
• /
• 2014

This paper proposed a discrete wavelet transform based method for time-varying physical parameter identification of shear type structures. The time-varying physical parameters are dispersed and expanded at multi-scale as profile and detail signal using discrete wavelet basis. To reduce the number of unknown quantity, the wavelet coefficients that reflect the detail signal are ignored by setting as zero value. Consequently, the time-varying parameter can be approximately estimated only using the scale coefficients that reflect the profile signal, and the identification task is transformed to an equivalent time-invariant scale coefficient estimation. The time-invariant scale coefficients can be simply estimated using regular least-squares methods, and then the original time-varying physical parameters can be reconstructed by using the identified time-invariant scale coefficients. To reduce the influence of the ill-posed problem of equation resolving caused by noise, the Tikhonov regularization method instead of regular least-squares method is used in the paper to estimate the scale coefficients. A two-story shear type frame structure with time-varying stiffness and damping are simulated to validate the effectiveness and accuracy of the proposed method. It is demonstrated that the identified time-varying stiffness is with a good accuracy, while the identified damping is sensitive to noise.

• Design & Manufacturing
• /
• 제16권3호
• /
• pp.1-8
• /
• 2022

In this study, an artificial neural network(ANN) was constructed to establish the relationship between process condition prameters and the qualities of the injection-molded product in the injection molding process. Six process parmeters were set as input parameter for ANN: melt temperature, mold temperature, injection speed, packing pressure, packing time, and cooling time. As output parameters, the mass, nominal diameter, and height of the injection-molded product were set. Two learning structures were applied to the ANN. The single-task learning, in which all output parameters are learned in correlation with each other, and the multi-task learning structure in which each output parameters is individually learned according to the characteristics, were constructed. As a result of constructing an artificial neural network with two learning structures and evaluating the prediction performance, it was confirmed that the predicted value of the ANN to which the multi-task learning structure was applied had a low RMSE compared with the single-task learning structure. In addition, when comparing the quality specifications of injection molded products with the prediction values of the ANN, it was confirmed that the ANN of the multi-task learning structure satisfies the quality specifications for all of the mass, diameter, and height.

• 한국전자파학회논문지
• /
• 제28권8호
• /
• pp.619-627
• /
• 2017

MIMO(Multi-Input Multi-Output) 레이다는 최근 여러 장점으로 각광을 받고 있는 새로운 개념의 레이다 기술로 다양한 신호처리 기술이 연구되고 있다. 그러나 전자전의 관점에서 보면 MIMO 레이다는 기존 레이다와 다른 동작원리를 가지고 있으므로 기존 레이다 파라미터 탐지 기술이 적용되지 않기에 새로운 파라미터 탐지 기술이 MIMO 레이다의 무력화와 효과적인 기만을 위하여 요구된다. 본 논문에서는 ULA(Uniform Linear Array) MIMO 레이다의 중요 파라미터인 직교신호의 개수를 2개의 이동 전파탐지기를 이용하여 암맹적으로 추정하거나 저잡음 상황에서 최소 3개의 전파탐지기를 이용하여 암맹적으로 추정하는 두 가지 기법을 제안하고 시뮬레이션을 통하여 성능을 확인한다.

• 한국전자통신학회논문지
• /
• 제16권1호
• /
• pp.53-60
• /
• 2021

고전압 CV 케이블은 편의성과 도시적 미관으로 인해 지하에 널리 설치되고 있습니다. 그러나 케이블 사고는 조립불량과 자연 열화로 인해 자주 발생하고 있다. 본 논문에서는 고전압 케이블 접속부의 최적 진단을 위한 다변수 측정 방법을 제안하고 그 기술적 유용성을 검증한다. 다변수 측정은 부분방전과 함께 진동과 열화상을 동시에 사용하여 케이블 연결 부품의 성능 저하를 진단하기 위한 기술이다. 다변수 측정의 유용성을 확인하기 위해 차폐 실험실에서 가속열화 실험을 수행하였고, 실험은 케이블 접속부의 열화를 12가지 유형으로 정의하여 각 열화 샘플을 제작하였다. 실험 결과, 진동신호 방법은 열화 진행과의 상관관계가 있는 결함 종류를 확인할 수 있었다. 열화상 검사의 경우 일부 결함의 진행과 관련하여 일관성이 발견되었다. 본 논문에서 제안된 방법은 현장에서 유용할 것으로 판단된다.