• 한국정밀공학회지
• /
• 제20권10호
• /
• pp.169-176
• /
• 2003

Genetic algorithm (GA), which has a powerful searching ability and is comparatively easy to use and also to apply, is in the spotlight in the field of the optimization for mechanical systems these days. However, it also contains some problems of slow convergence and low efficiency caused by a huge amount of repetitive computation. To improve the processing efficiency of repetitive computation, some papers have proposed paralleled GA these days. There are some cases that mention the use of gray code or suggest using gray code partially in GA to raise its slow convergence. Gray code is an encoding of numbers so that adjacent numbers have a single digit differing by 1. A binary gray code with n digits corresponds to a hamiltonian path on an n-dimensional hypercube (including direction reversals). The term gray code is open used to refer to a reflected code, or more specifically still, the binary reflected gray code. However, according to proposed reports, gray code GA has lower convergence about 10-20% comparing with binary code GA without presenting any results. This study proposes new Full gray code GA (FGGA) applying a gray code throughout all basic operation fields of GA, which has a good data processing ability to improve the slow convergence of binary code GA.

• Nuclear Engineering and Technology
• /
• 제55권6호
• /
• pp.2215-2221
• /
• 2023

In order to further meet the requirements of weight, volume, and dose minimization for new nuclear energy devices, the bare-bones multi-objective particle swarm optimization algorithm is used to automatically and iteratively optimize the design parameters of radiation shielding system material, thickness, and structure. The radiation shielding optimization program based on the bare-bones particle swarm optimization algorithm is developed and coupled into the reactor radiation shielding multi-objective intelligent optimization platform, and the code is verified by using the Savannah benchmark model. The material type and thickness of Savannah model were optimized by using the BBMOPSO algorithm to call the dose calculation code, the integrated optimized data showed that the weight decreased by 78.77%, the volume decreased by 23.10% and the dose rate decreased by 72.41% compared with the initial solution. The results show that the method can get the best radiation shielding solution that meets a lot of different goals. This shows that the method is both effective and feasible, and it makes up for the lack of manual optimization.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2010년도 춘계학술대회 논문집
• /
• pp.13-14
• /
• 2010

• 유체기계공업학회:학술대회논문집
• /
• 유체기계공업학회 2002년도 유체기계 연구개발 발표회 논문집
• /
• pp.47-53
• /
• 2002

The hydraulic design optimization and performance analysis of mixed-flow pumps for waterjet marine vehicle propulsion has been carried out using mean streamline analysis and three-dimensional computational fluid dynamics (CFD) code. In the present study the conceptual design optimization has been formulated with a non-linear objective function to minimize the fluid dynamic losses and then the commercial CFD code was incorporated to allow for detailed flow dynamic phenomena in the pump system. New designed mixed-flow model pump has been tested in the laboratory. Predicted performance curves by the CFD code agree very well with experimental data for a newly designed mixed-flow pump over the normal operating conditions. The design and prediction methods presented herein can be used efficiently as a unified hydraulic design process of mixed-flow pumps for waterjet marine vehicle propulsion.

• 한국멀티미디어학회논문지
• /
• 제9권9호
• /
• pp.1247-1256
• /
• 2006

가상 기계란 하드웨어로 이루어진 물리적 시스템과는 달리 소프트웨어로 제작되어 논리적인 시스템 구성을 갖는 개념적인 컴퓨터이다. 그러나 가상 기계는 실제 프로세서로 처리하는 것보다 실행 속도가 매우 느리기 때문에 실행되는 코드의 최적화가 매우 중요하다. 본 논문은 가상 기계 코드 최적화기의 실험대상으로 EVM(Embedded Virtual Machine)의 중간 코드인 SIL(Standard Intermediate Language)을 이용하였다. 현존하는 최적화 방법론에 관한 연구를 통하여 가상 기계 코드 특성을 고려한 최적화 방법론을 제시하고, 최적화된 코드를 생성하기 위한 코드 최적화기를 설계하고 구현하였다. 가상 기계 코드 최적화기는 주어진 패턴을 찾아서 패턴에 해당하는 부분을 최적화 코드로 바꾸어, 전체 코드의 크기를 줄이고 실행 속도의 개선효과를 가진다. 또한, 구현된 최적화기의 실험 결과를 도출하였다.

• 한국정보처리학회논문지
• /
• 제6권8호
• /
• pp.2253-2261
• /
• 1999

ACK에서는 패턴 테이블 생성기와 핍홀 최적화기에서 스트링 패턴 매칭 기법을 이용하여 EM 중간 코드에 대한 최적화 코드를 생성한다. 하지만 이 스트링 패턴 매칭 방법은 패턴 결정 시에 반복적으로 많은 비교 동작이 이루어지므로 비효율적이다. 본 논문은 ACK의 중간 코드 최적화기를 개선하기 위해 EM 트리 생성기, 최적화 패턴 테이블 생성기, 트리 패턴 매칭기로 구성된 트리 패턴 매칭 알고리즘을 이용한 EM 중간 코드 최적화 시스템을 설계하고 구현하였다. 이러한 트리 패턴 매칭 알고리즘은 EM 트리를 하향식으로 순회하면서 트리 구조를 가진 패턴 테이블을 참조하여 루트 노드를 중심으로 패턴 매칭을 수행한다. 트리 패턴 매칭 동작은 궁극적으로 ACK의 스트링 패턴 매칭에 비해 최적화 패턴을 찾는데 걸리는 시간을 평균 10.8% 감소시킬 수 있는 효과를 보였다.

• 한국항공운항학회지
• /
• 제14권3호
• /
• pp.7-15
• /
• 2006

Aerodynamic analysis of an airfoil in the transonic region was automated in order to enable parametric study by using the journal file of the commercial analysis code FLUENT, pre/post process Gambit and computational mathematics code MATLAB. The automated capability was illustrated via NACA 0012 and RAE 2822 airfoils. This analysis was carried out at Mach numbers ranged from 0.70 to 0.80, angles of attack; 1$^{\circ}$, 2$^{\circ}$ and 4$^{\circ}$, Reynolds numbers; 4.0${\times}$106, 6.5${\times}$106. The analysis results of a pressure coefficient were verified by comparing with the experimental data which were measured in terms of chord length because the pressure coefficient of an airfoil surface is a good estimator of flow characteristics. The results of two airfoils show that this analysis code is useful enough to be used in the design optimization of airfoil.

• Nuclear Engineering and Technology
• /
• 제56권2호
• /
• pp.520-525
• /
• 2024

The optimization design problem of nuclear reactor radiation shield is a typical multi-objective optimization problem with almost 10 sub-objectives and the sub-objectives are always demanded to be under tolerable limits. In this paper, a design method combining multi-objective optimization algorithms with paralleling discrete ordinate transportation code is developed and applied to shield design of the Savannah nuclear reactor. Three approaches are studied for light-weighted and compact design of radiation shield. Comparing with directly optimization with 10 objectives and the single-objective optimization, the approach by setting sub-objectives representing weight and volume as optimization objectives while treating other sub-objectives as constraints has the best performance, which is more suitable to reactor shield design.

• 대한기계학회논문집B
• /
• 제26권6호
• /
• pp.795-804
• /
• 2002

The optimal design code of an axial flow pump has been developed to determine geometric and fluid dynamic variables under hydrodynamic as well as mechanical design constraints. The design code includes the optimization of the complete radial distribution of the geometry by determining the coefficients of 2$^{nd}$ order polynomials to represent the three-dimensional geometry. The optimization problem has been formulated with a nonlinear multivariable objective function, maximizing the efficiency and stall margin, while minimizing the net positive suction head required. Calculation of the objective function is based on the mean streamline analysis and through-flow analysis using the present state-of-the-art model. The optimal solution is calculated using the penalty function method in which the genetic optimizer is employed. The optimized efficiency and design variables are presented in this paper as a function of non-dimensional specific speed in the range, 2$\leq$ $n_{s}$ $\leq$10. The results can be used in preliminary design of axial flow pumps.

• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2001년도 ICCAS
• /
• pp.169.4-169
• /
• 2001

Conventional methods to solve the nonlinear programming problem range from augmented Lagrangian methods to sequential quadratic programming (SQP) methods. NPSOL, which is a SQP code, has been widely used to solve various optimization problems but is still subject to many numerical problems such as convergence to local optima, difficulties in initialization and in handling non-smooth cost functions. Recently, many evolutionary methods have been developed for constrained optimization. Among them, CEALM (Co-Evolutionary Augmented Lagrangian Method) shows excellent performance in the following aspects: global optimization capability, low sensitivity to the initial parameter guessing, and excellent constraint handling capability due to the benefit of the augmented Lagrangian function. This algorithm is ...