• 한국공작기계학회논문집
• /
• 제16권5호
• /
• pp.205-210
• /
• 2007

The integration method of binary and real encoding in genetic algorithm is proposed to deal with design variables of various types in gear drive design. The method is applied to optimum design of multi-stage gear drive. Integer and Discrete type design variables represent the number of teeth and module, and continuous type design variables represent face width, helix angle and addendum modification factor etc. The proposed genetic algorithm is applied for the gear ratio optimization and the volume optimization(minimization) of multi-stage geared motor which is used in field. In result, the proposed design optimization method shows an effectiveness in optimum design process and the new design has a better results compared with the existing design.

• 대한조선학회논문집
• /
• 제31권4호
• /
• pp.147-156
• /
• 1994

선체의 구조설계는 최적화 방법을 이용하여 상당히 오래 전부터 최적 구조설계 방법을 사용해 오고 있었으나, 대부분의 경우, 설계변수(設計變數)를 연속적인 실수(實數)로 가정하여 최적해를 구하거나, 아니면 실수(實數)와 정수(整數)가 혼합된 문제에 대해서는 뚜렷한 해결 방안을 제시하지 못하고 있는 실정이다. 특히 최적해의 국부(局部) 최적성 내지는 이산적(離散的) 변수 특성이 있는 최적설계 문제에 대해서는 몇개의 초기치를 사용하여 얻어진 최적해를 상호 비교하여 주어진 문제의 전체적(全體的) 최적해를 구하고자 하였다. 많은 경우 이러한 방법은 확실한 대안이 되지 못하고 본질적인 문제점은 미해결로서 남아 있어 왔다. 그래서 본 연구에서는 생물의 진화 법칙을 모사한 유전적(遺傳的) 알고리즘을 이용하여 선체 구조물의 최적설계시 고려해야 하는 보강재의 갯수를 정수(整數)로 취급하는 문제라든지 판 두께와 같이 이산적(離散的) 특성을 갖는 설계변수 문제 등(等)이 최적설계에 미치는 영향을 검토하여 보다 일반적인 최적화 방법으로서 유전적(遺傳的) 알고리즘의 유용성을 확인하였다.

• Journal of Electrical Engineering and Technology
• /
• 제4권2호
• /
• pp.185-193
• /
• 2009

This paper presents a methodology for photovoltaic (PV) system allocation in distribution systems using a discrete particle swarm optimization (DPSO). The PV allocation problem is in the category of mixed integer nonlinear programming and its formulation may include multi-valued dis-crete variables. Thus, the PSO requires a scheme to deal with multi-valued discrete variables. This paper introduces a novel multi-level quantization scheme using a sigmoid function for discrete particle swarm optimization. The technique is employed to a standard PSO architecture; the same velocity update equation as in continuous versions of PSO is used but the particle's positions are updated in an alternative manner. The set of multi-level quantization is defined as integer multiples of powers-of-two terms to efficiently approximate the sigmoid function in transforming a particle's position into discrete values. A comparison with a genetic algorithm (GA) is performed to verify the quality of the solutions obtained.

• 전산구조공학
• /
• 제10권2호
• /
• pp.255-263
• /
• 1997

실제로 구조시스템들의 최적설계는 설계변수가 연속값이 아닌 이산값을 요하는 경우가 대부분이다. 본 논문은 이산형 설계변수를 갖는 비대칭 복합 적층평판에 대해 선형 근사화방법을 이용한 이산최적설계를 수행하였으며, 이 방법이 매우 효율적임을 보였다. 대상 문제는 축력, 전단력, 그리고 휨과 비틀림 모멘트의 평면 내하중들(in-plane loads)의 다중하중조건을 받는 것으로 고려하였으며, 복합 적층평판을 구성하는 플라이들에 대한 최대변형률 규준을 설계 제약조건으로 부과하였다. 이산 최적화를 위한 초기 접근방법으로 단 한번의 연속변수 최적화 과정이 FDM(Feasible Direction Method)을 이용하여 수행되었으며, 차후 이산 및 연속변수를 포함하는 비선형 이산최적화문제를 SLDP(Sequential Linear Discrete Programming)방법에 의해 선형 근사화된 혼합정수계획문제로 형성하여 풀었다. 수치예에서 6개의 플라이로 구성된 비대칭 복합 적층평판을 대상으로 회전식 적층배열([(90-.theta.)/-(60+.theta.)/-.theta./-(45+.theta.)/(45-.theta.)]/sub s/)에 따른 이산최적해를 구하였다. 효율성 입증을 위해 똑같은 문제를 비선형 분기한계법을 이용하여 풀었으며, 그 결과를 비교 분석하였다.

• International Journal of Precision Engineering and Manufacturing
• /
• 제1권1호
• /
• pp.62-70
• /
• 2000

The design of gear train is a kind of mixed problems which have to determine various types of design variables; i,e., continuous, discrete, and integer variables. Therefore, the most common practice of optimum design using the derivative of objective function has difficulty in solving those kinds of problems and the optimum solution also depends on initial guess because there are many sophisticated constrains. In this study, the Genetic Algorithm is introduced for the optimum design of gear trains to solve such problems and we propose a genetic algorithm based gear design system. This system is applied for the geometrical volume(size) minimization problem of the two-stage gear train and the simple planetary gear train to show that genetic algorithm is better than the conventional algorithm solving the problems that have continuous, discrete, and integer variables. In this system, each design factor such as strength, durability, interference, contact ratio, etc. is considered on the basis of AGMA standards to satisfy the required design specification and the performance with minimizing the geometrical volume(size) of gear trains

• 한국공작기계학회:학술대회논문집
• /
• 한국공작기계학회 2004년도 춘계학술대회 논문집
• /
• pp.118-123
• /
• 2004

In this study, genetic algorithm mixed binary and real encoding is proposed to deal with design variables of various types. And that is applied to optimum design of Multi-stage gear drive. Design of pressure vessel which is mixed discrete and continuous variables is applied to verify reasonableness of proposed genetic algorithm. The proposed genetic algorithm is applied for the gear ratio optimization and the volume minimization of geared motor which is used in field. In result, it shows that the volume has decreased about 8% compared with the existing geared motor.

• 전산구조공학
• /
• 제10권4호
• /
• pp.349-358
• /
• 1997

본 논문은 복합 적층구조의 최적설계에 있어서 유전알고리즘(GA)의 응용성을 보여준다. 설계점들의 최기집단이 확률론적 과정에 의해 무작위로 생성되고, 설계점들의 개선을 위해 자연선택과 적자생존의 원리가 적용되었다. 유전알고리즘의 범용성 및 신뢰성 검증을 위해 5가지 검증 함수를 고려하였으며, 수치예에서 연속형 및 정수형 그리고 이산형 설계변수를 동시에 갖는 복합 적층 캔틸레버보의 최소 중량 설계가 외부 벌칙함수가 부가된 유전알고리즘에 의해 수행되었다. 설계 문제는 강도, 변위 그리고 고유진동수 제약조건을 포함하면서 다차 비선형성으로 정식화 되었다. 수치예의 결과에 대한 비교분석을 통해 유전알고리즘 탐색 기법이 높은 범용성을 지니면서 양질의 최적해를 매우 효과적으로 찾게됨을 보였다.

• 한국안전학회지
• /
• 제14권2호
• /
• pp.70-76
• /
• 1999

In this paper, a mixed-integer programming approach is presented for adjusting the voltage profiles in a power system. The advent of large-scaled system makes the reactive power and voltage problem-an attempt to achieve an overall improvement of system security, service quality and economy-more complex and seriously, Although the problem is originally a nonlinear optimization problem, it can be formulated as a mixed integer linear programming(MILP) problem without deteriorating of solution accuracy to a certain extent. The MILP code is developed by the branch and bound process search for the optimal solution. The variable for modeling transformer tap positions is handled as discrete one, and other variables continuous ones. Numerical data resulting from case study using a modified IEEE 30 bus system with outaged line show that the MILP can produce more reductions of magnitude in the operating cost. The convergence characteristics of the results are also presented and discussed.

• 경영과학
• /
• 제13권1호
• /
• pp.97-109
• /
• 1996

Genetic Algorithms have been successfully applied to various problems (for example, engineering design problems with a mix of continuous, integer and discrete design variables) that could not have been readily solved with traditional computational techniques. But, several problems for which conventional Genetic Algorithms are ill defined are premature convergence of solution and application of exterior penalty function. Therefore, we developed an Improved Genetic Algorithms (IGAs) to solve above two problems. As a case study, IGAs is applied to several nonlinear optimization problems and it is proved that this algorithm is very useful and efficient in comparison with traditional methods and conventional Genetic Algorithm.

• 한국경영과학회지
• /
• 제13권1호
• /
• pp.97-97
• /
• 1988

Genetic Algorithms have been successfully applied to various problems (for example, engineering design problems with a mix of continuous, integer and discrete design variables) that could not have been readily solved with traditional computational techniques. But, several problems for which conventional Genetic Algorithms are ill defined are premature convergence of solution and application of exterior penalty function. Therefore, we developed an Improved Genetic Algorithms (IGAs) to solve above two problems. As a case study, IGAs is applied to several nonlinear optimization problems and it is proved that this algorithm is very useful and efficient in comparison with traditional methods and conventional Genetic Algorithm.