개선된 유전 알고리즘의 최적해 비교에 관한 연구 = A Study on The Optimal Solution Comparison Of Revised Genetic Algorithms

다국어 초록 (Multilingual Abstract)

The Optimization problem is to select the best of many possible design alternatives in a complex design space. Genetic Algorithms, one of numerous techniques to search optimal solution, have been successfully applied to various problem(for example, parameter tuning in expert system, structural systems with a mix of continuous integer and discrete design variables) that could not have been readily solved with more conventional computational techniques, But, Conventional Genetic Algorithms are ill defined for two classes of problems ie, parameter convergence and penalty function.
Therefore, this paper developed Revised Algorithms to solve these problems, As a case study, sin function example and design of gear train are demonstrated to show the effectiveness of The two Revised Genetic Algorithms(R. G. A & H. G. A.) and one of the Revised Genetic Algorithms was more effectiveness. That's H. G. A.
It has robustness and problem-independent characteristics. In the reproduction step of the H. G. A. the roulette wheel method is replaced the random search and the creeping value method is modified to improve local optimum solutions in the mutation step. The encoding and decoding procedures are replaced with the real number representation method in the H. G. A.

목차 (Table of Contents)

• 1. 서론
• 2. Genetic Algorithm 일반적 고찰
• 3. Genetic Algorithm의 비교/분석
• 4. 기어 트레인의 설계
• 5. 결론
• 참고문헌