• 대한조선학회논문집
• /
• 제46권6호
• /
• pp.562-568
• /
• 2009

Hull form generation and variation methods to be mainly discussed in this study are based on the fairness optimized B-Spline form parameter curves (FOBFC). These curves can be used both as indirect modification function for variation and as geometric entities for hull form generation. The flexibility and functionality of geometric control technique play the most important role for the success of hull form optimization. This study shows the hydrodynamic optimization process and the characteristics of optimum design hull forms of a 14,000TEU containership and 60K LPG carrier. SHIPFLOW has been used as a CFD solver and FS-Framework as a geometric modeler and optimizer.

• 스마트미디어저널
• /
• 제9권4호
• /
• pp.36-43
• /
• 2020

The Monte Carlo tree search (MCTS) is a popular method for implementing an intelligent game program. It has several hyper-parameters that require an optimization for showing the best performance. Due to the stochastic nature of the MCTS, the hyper-parameter optimization is difficult to solve. This paper uses the self-playing capability of the MCTS-based game program for optimizing the hyper-parameters. It seeks a winner path over the hyper-parameter space while performing the self-play. The top-q longest winners in the winner path compete for the final winner. The experiment using the 15-15-5 game (Omok in Korean name) showed a promising result.

• Structural Engineering and Mechanics
• /
• 제71권1호
• /
• pp.45-56
• /
• 2019

A modified mechanical model of pre-pressed spring self-centering energy dissipation (PS-SCED) brace is proposed, and the hysteresis band is distinguished by the indication of relevant state variables. The MDOF frame system equipped with the braces is formulated in an incremental form of linear acceleration method. A multi-objective genetic algorithm (GA) based brace parameter optimization method is developed to obtain an optimal solution from the primary design scheme. Parameter sensitivities derived by the direct differentiation method are used to modify the change rate of parameters in the GA operator. A case study is conducted on a steel braced frame to illustrate the effect of brace parameters on node displacements, and validate the feasibility of the modified mechanical model. The optimization results and computational process information are compared among three cases of different strategies of parameter change as well. The accuracy is also verified by the calculation results of finite element model. This work can help the applications of PS-SCED brace optimization related to parameter sensitivity, and fulfill the systematic design procedure of PS-SCED brace-structure system with completed and prospective consequences.

• Structural Engineering and Mechanics
• /
• 제64권5호
• /
• pp.567-577
• /
• 2017

The traditional design procedure of a prestressed concrete (PC) cable-stayed bridge is complex and time-consuming. The designers have to repeatedly modify the configuration of the large number of design parameters to obtain a feasible design scheme which maybe not an economical design. In order to efficiently achieve an optimum design for PC cable-stayed bridges, a multi-parameter optimization technique is proposed. In this optimization technique, the number of prestressing tendons in girder is firstly set as one of design variables, as well as cable forces, cable areas and cross-section sizes of the girders and the towers. The stress and displacement constraints are simultaneously utilized to ensure the safety and serviceability of the structure. The target is to obtain the minimum cost design for a PC cable-stayed bridge. Finally, this optimization technique is carried out by a developed PC cable-stayed bridge optimization program involving the interaction of the parameterized automatically modeling program, the finite element structural analysis program and the optimization algorithm. A low-pylon PC cable-stayed bridge is selected as the example to test the proposed optimization technique. The optimum result verifies the capability and efficiency of the optimization technique, and the significance to optimize the number of prestressing tendons in the girder. The optimum design scheme obtained by the application can achieve a 24.03% reduction in cost, compared with the initial design.

• 한국통신학회논문지
• /
• 제41권8호
• /
• pp.917-920
• /
• 2016

전통적 도래각 추정기법[1]과 별개로 2004년 이후 입사신호의 입사방향은 공간 영역에서 희소도(sparsity)를 가짐을 이용한 도래각 추정 기법이 제안되었다. 압축센싱 기반 도래각 추정 알고리즘인 SpSF 알고리즘에 이용되는 비용함수는 비선형 다변수 최적화문제이다. 적절한 변환을 통하여 해당 비용함수는 볼록 최적화 (convex optimization) 문제로 표현할 수 있다. 볼록 최적화 문제는 제한조건이 있는 최적화 문제이며 제한조건에 포함되는 상수를 지정해야 한다. 본 연구에서는 제한조건에 포함되는 사용자지정 상수값 결정법을 제안한다. 잡음의 실수부와 허수부가 서로 독립인 평균 0인 정규분포를 따름을 이용하여 제한조건에 포함되는 행렬의 Frobenius norm의 평균을 유도할 수 있으며, 이를 이용하여 제한조건에 포함되는 상수를 결정할 수 있다. 제안된 방법에 의해 결정된 상수를 이용한 SpSF 알고리즘이 실제로 동작함을 보였다.

• 전기전자학회논문지
• /
• 제24권4호
• /
• pp.969-974
• /
• 2020

Research into hyper parameter optimization (HPO) has recently revived with interest in models containing many hyper parameters, such as deep neural networks. In this paper, we introduce the most widely used HPO methods, such as grid search, random search, and Bayesian optimization, and investigate their characteristics through experiments. The MNIST data set is used to compare results in experiments to find the best method that can be used to achieve higher accuracy in a relatively short time simulation. The learning rate and weight decay have been chosen for this experiment because these are the commonly used parameters in this kind of experiment.

• Journal of the Korean Data and Information Science Society
• /
• 제12권1호
• /
• pp.11-18
• /
• 2001

다구찌가 제안한 파라미터 설계는 제품설계나 공정설계 단계에서 품질특성치의 수행변동(performance variation)을 줄이는데 있다. 파라미터 설계에서 품질평균 근처에서 품질의 변동을 줄일 수 있는 최적조건을 찾는데 있어서 신호대 잡음비(Signal-to-noise ratio; SN비)라는 수행측도(performance measure)를 사용하였다. 그러나 많은 통계학자들은 SN비를 이용한 다구찌 분석 기법을 비판한다. 본 논문에서는 파라미터 설계를 위한 최적조건을 찾는데 있어서 SN비를 사용하지 않는 보다 실질적인 최적방안을 제안한다.

• Advances in Automotive Engineering
• /
• 제1권1호
• /
• pp.1-20
• /
• 2018

The aim of this study is the optimization of the parameters of interconnected Hydro-Pneumatic (HP) suspension system of a three-axle vehicle for ride comfort and handling. For HP suspension systems of equivalent vertical stiffness and damping characteristics, interconnected HP suspension systems increase roll and pitch stiffness and damping characteristics of the vehicle as compared to unconnected HP suspension systems. Thus, they result in improved handling and braking/acceleration performances of the vehicle. However, increased roll and pitch stiffness and damping characteristics also increase roll and pitch accelerations, which in turn result in degraded ride comfort performance. Therefore, in order to improve both ride comfort and vehicle handling performances simultaneously, an optimum parameter set of an interconnected HP suspension system is obtained through an optimization procedure. The objective function is formed as the sum of the weighted vertical accelerations according to ISO 2631. The roll angle, one of the important measures of vehicle handling and driving safety, is imposed as a constraint in the optimization study. Upper and lower parameter bounds are used in the optimization in order to get a physically realizable parameter set. Optimization procedure is implemented for a three-axle vehicle with unconnected and interconnected suspension systems separately. Optimization results show that interconnected HP suspension system results in improvements in both ride comfort and vehicle handling performance, as compared to the unconnected suspension system. As a result, interconnected HP suspension systems present a solution to the conflict between ride comfort and vehicle handling which is present in unconnected suspension systems.

• 한국멀티미디어학회논문지
• /
• 제6권1호
• /
• pp.136-147
• /
• 2003

주어진 문제에 대하여 최적의 성능을 가지는 신경회로망을 얻기 위해서는 학습을 통한 매개변수의 최적화 (parameter optimization)와 모델 선택을 통한 구조 최적화(structure optimization )의 통합적인 과정이 필요하다. 본 논문에서는, 각 세부 방법들의 특성을 고려하여, 공통의 특성을 갖는 방법들을 결합함으로써 효율적이면서도 일반화 성능을 높이는 총체적인 신경회로망 최적화 방법을 제안한다. 먼저 다양한 오차 함수를 사용할 수 있는 자연 기울기 강하 학습에 적응적 정규화 방법을 도입함으로써 가중치 매개변수(weight parameter)들을 최적화한다. 그리고 이렇게 최적화된 매개변수(parameter)들에 자연 프루닝(natural pruning)을 적용하여 불필요한 요소들을 제저하여 최적화 된 구조를 생성한다. 반복적인 과정에 의하여 후보 모델들을 구성하고 베이시안 정보 기준(Bayesian Information Criterion: BIC )을 이 용하여 최적의 모델을 평가하여 선택하는 방법을 제안하였다. 벤치마크 데이터에 대한 실험을 통하여 제안하는 방법의 구조 최적화 능력과 일반화 성능의 우수성을 보였다.

• 대한전기학회논문지
• /
• 제39권8호
• /
• pp.792-804
• /
• 1990

This paper describes an efficient optimization algorithm by calculating sensitivity function for power system stabilization. In power system, the dynamic performance of exciter, governor etc. following a disturbance can be presented by a nonlinear differential equation. Since a nonlinear equation can be linearized for small disturbances, the state equation is expressed by a system matrix with system parameters. The objective function for power system operation will be related to the system parameter and the initial state at the optimal control condition for control or stabilization. The object function sensitivity to the system parameter can be considered to be effective in selecting the optimal parameter of the system.