• 한국해양공학회지
• /
• 제17권6호
• /
• pp.38-46
• /
• 2003

Aim of this article is to propose Micro-Genetic Simulated Annealing (${\mu}GSA$) as a hybrid metaheuristics approach to find the global optimum of nonlinear optimisation problems. This approach combines the features of modern metaheuristics such as micro-genetic algorithm (${\mu}GAs$) and simulated annealing (SA) with the general robustness of parallel exploration and asymptotic convergence, respectively. Therefore, ${\mu}GSA$ approach can help in avoiding the premature convergence and can search for better global solution, because of its wide spread applicability, global perspective and inherent parallelism. For the superior performance of the ${\mu}GSA$, the five well-know benchmark test functions that were tested and compared with the two global optimisation approaches: scatter search (SS) and hybrid scatter genetic tabu (HSGT) approach. A practical application to structural sandwich panel is also examined by optimism the weight function. From the simulation results, it has been concluded that the proposed ${\mu}GSA$ approach is an effective optimisation tool for soloing continuous nonlinear global optimisation problems in suitable computational time frame.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2003년도 가을 학술발표회 논문집
• /
• pp.115-122
• /
• 2003

A Metropolis genetic algorithm(MGA) is developed and applied for the structural design optimization. In MGA favorable features of Metropolis algorithm in simulated annealing(SA) are incorporated in simple genetic algorithm(SGA), so that the MGA alleviates the disadvantage of finding imprecise solution in SGA and time-consuming computation in SA. Performances of MGA are compared with those of conventional algorithms such as Holland's SGA, Krishnakumar's micro genetic algorithm(μGA), and Kirkpatrick's SA. Typical numerical examples are used to evaluate the favorable features and applicability of MGA From the theoretical evaluation and numerical experience, it is concluded that the proposed MGA is a reliable and efficient tool for structural design optimization.

• 한국전산구조공학회논문집
• /
• 제19권1호
• /
• pp.27-37
• /
• 2006

모사풀림(SA)의 특징적인 Metropolis 규준을 단순 유전자 알고리즘(SGA)의 재생산 연산과정에 도입함으로써 Metropolis 유전자 알고리즘(MGA)이 개발되고, 구조 설계 최적화에 응용되었다. 이러한 결합을 통하여 MGA는 개체의 다양성을 유지하며, 초기 세대에서 나타날 수 있는 유용한 유전자 정보가 보존될 수 있다. 따라서 이 연구에서 제안된 MGA는, 국부적 최적해로 조기 수렴하게 되는 SGA의 단점과 정밀한 전역적 최적해를 찾기 위해 수없이 많은 계산을 해야 하는 SA의 단점을 극복하게 되었다 수치예를 통하여 MGA의 성능과 적용성을 재래의 알고리즘들과 비교하고 평가하였다. 특히 MGA의 성능 신뢰성과 강건성을 평가하는 데는 집단 크기와 최대 반복세대수의 효과를 인용하였다. 이론적 고찰과 수치예의 결과로부터, 이 연구에서 개발된 MGA가 효율성과 신뢰성을 갖춘 구조 설계 최적화의 도구로서 평가되었다.

• 한국전산구조공학회:학술대회논문집
• /
• 한국전산구조공학회 2008년도 정기 학술대회
• /
• pp.416-421
• /
• 2008

A Metropolis genetic algorithm (MGA) is a newly-developed hybrid algorithm combining simple genetic algorithm (SGA) and simulated annealing (SA). In the algorithm, favorable features of Metropolis criterion of SA are incorporated in the reproduction operations of SGA. This way, MGA alleviates the disadvantages of finding imprecise solution in SGA and time-consuming computation in SA. It has been successfully applied and the efficiency has been verified for the practical structural design optimization. However, applicability of MGA for the wider range of problems should be rigorously proved through the solution of mathematical optimization problems. Thus, performances of MGA for the typical mathematical problems are investigated and compared with those of conventional algorithms such as SGA, micro genetic algorithm (${\mu}GA$), and SA. And, for better application of MGA, the effects of acceptance level are also presented. From numerical Study, it is again verified that MGA is more efficient and robust than SA, SGA and ${\mu}GA$ in the solution of mathematical optimization problems having various features.