• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.36C no.3
• /
• pp.58-64
• /
• 1999

Evolutionary programming is mainly characterized by two factors; one is the selection strategy and the other the mutation rule. In this paper, a new mutation rule that is the same form of well-known backpropagation learning rule of neural networks has been presented. The proposed mutation rule adapts the best individual's value as the target value at the generation. The temporal error improves the exploration through guiding the direction of evolution and the momentum speeds up convergence. The efficiency and robustness of the proposed algorithm have been verified through benchmark test functions.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.35C no.9
• /
• pp.81-92
• /
• 1998

A new hybrid technique using several sub-populations having completely different evolutionary behaviors is proposed to increase the possibility to quickly find the global optimum of continuous optimization problem. It has three sub-populations. Two NPOSA algorithms showing good performance in the problem having a rugged fitness function are applied to two sub-populations and a self-adaptive evolutionary algorithm to the other sub-population. Sub-populations evolve in different manners and the interaction among these sub-populations lead to the global optimum quickly. The efficiency of this technique is verified through benchmark test functions. Finally, the algorithm with three sub-populations has been applied to seek for the optimal camera calibration parameters. After an error function has been defined using measured feature points of a calibration block, it has been shown that the algorithm searches for the camera parameters that minimize the error function.

• Proceedings of the IEEK Conference
• /
• 2003.07d
• /
• pp.1247-1250
• /
• 2003

본 논문에서는 다수의 개념을 효과적으로 학습할 수 있는 논리적인 진화방식을, 보다 일반적인 문제에 적용하기 위하여 확장하였다. 기존의 알고리듬이 0 차 논리(zero-order logic)에 의해 표현되는 개념만을 학습할 수 있었던 점을 개선하여, 특징논리로 표현되는 개념까지 학습할 수 있도록 확장하였다. 또한 확장된 알고리듬을 이용하여, 실수입력을 가지는 개념 또한 학습할 수 있도록 하였다. 제안된 알고리듬은 이전의 알고리듬과 마찬가지로, 다수의 개념을 학습할 때, 이전의 학습결과를 이용하여 새로운 개념을 쉽게 학습할 수 있는 특징을 가지고 있다. 제안된 알고리듬에 의해 특징논리로 표현되는 개념을 학습할 수 있으며, 다수의 개념을 학습할 때 성능향상이 이루어지는 것을 실험을 통해 확인하였다.

• Korean Management Science Review
• /
• v.18 no.2
• /
• pp.117-124
• /
• 2001

This paper presents a machining parameter selection approach using an evolutionary computation (EC). In order to perform a successful material cutting process, the engineer is to select suitable machining parameters. Until now, it has been mostly done by the handbook look-up or solving optimization equations which is inconvenient when not in handy. The main thrust of the paper is to provide a handy machining parameter selection approach. The EC is applied to rapidly find optimal machining parameters for the user\\`s specific machining conditions. The EC is basically a combination of genetic a1gorithm and microcanonical stochastic simulated annealing method. The approach is described in detail with an application example. The paper concludes with a discussion on the potential of the proposed approach.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2006.05a
• /
• pp.1036-1041
• /
• 2006

본 연구에서는 허브와 스포크의 입지선정과 차량 경로가 통합된 hub-and-spoke 네트워크 설계문제를 다룬다. Hub-and-spoke 네트워크는 대량화와 공동화를 통해 물류효율화를 실현하기 위한 대표적인 구조로 물류시스템에서 흔히 사용되고 있다. 이러한 물류시스템에서 물류비용의 절감과 고객서비스 향상을 위한 효율적인 수송네트워크 설계는 매우 중요하다. 전통적인 hub-and-spoke 네트워크 설계문제에서 각 스포크의 위치와 화물량이 미리 주어진 상황에서 허브의 입지를 결정하였다. 하지만 스포크 역시 스포크가 담당하는 고객들의 위치와 담당 영역에 따라 그 위치와 수, 그리고 화물량이 변할 수 있다. 또한 정확한 비용산출을 위해서는 스포크에서 고객으로의 수집과 배달을 위한 차량경로가 함께 고려되어야 한다. 다루는 수송망 설계문제는 상호 관련성 있는 여러 부분문제가 결합된 통합문제로써 이를 해결하는 방법으로 기존의 발견적 방법에 의한 순차적 기법은 한계가 있다. 본 연구에서는 공생 진화알고리듬 기반의 방법론을 채용하여 다루는 수송망 설계문제를 동시에 통합적으로 해결할 수 있는 알고리듬을 개발한다. 실험을 통해 개발한 알고리듬의 우수성과 그 적용성을 보인다.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1998.06c
• /
• pp.71-74
• /
• 1998

HMM (hidden Markov model)을 이용한 음성인식은 현재 가장 널리 쓰여지고 있는 방법으로, 이 중 CDHMM (continuous observation density HMM)은 상태에서 관측심볼확률을 연속확률밀도를 사용하여 표현한다. 본 논문에서는 가우스 혼합밀도함수를 사용하는 CDHMM의 상태천이확률과, 관측심볼확률을 표현하기 위한 인자인 평균벡터, 공분산 행렬, 가지하중값을 유전자 알고리듬을 사용하여 최적화하는 방법을 제안하였다. 유전자 알고리듬은 매개변수 최적화문제에 대하여 자연의 진화원리를 모방한 알고리듬으로, 염색체 형태로 표현된 개체군 (population) 중에서 환경에 대한 적합도 (fitness)가 높은 개체가 높은 확률로 살아남아 재생 (reproduction)하게 되며, 교배 (crossover)와 돌연변이 (mutation) 연산 후에 다음 세대 개체군을 형성하게 되고, 이러한 과정을 반복하면서 최적의 개체를 구하게 된다. 본 논문에서는 상태천이확률, 평균벡터, 공분산행렬, 가지하중값을 부동소수점수 (floating point number)의 유전자형으로 표현하여 유전자 알고리듬을 수행하였다. 유전자 알고리듬은 복잡한 탐색공간에서 최적의 해를 찾는데 효과적으로 적용되었다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.7 no.3
• /
• pp.3-11
• /
• 1997

One of major research areas in the recurrent neural network is to develop stable learning algorithm. In this paper, the stable learning algorithm is developed by utilizing the evolutionary programming. The effectiveness of the proposed learning algorithm will be verified by simulating two d.0.f. robot manipulator.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.10 no.5
• /
• pp.432-441
• /
• 2000

This paper deals with the fuzzy modeling for the complex and uncertain nonlinear systems, in which conventional and mathematical models may fail to give satisfactory results. Genetic algorithm has been used to identifY parameters and structure of fuzzy model because it has the ability to search optimal solution somewhat globally. The genetic algorithm, however, has a problem, which optimization process can be premature convergence in the case of lack of genetic divergence of population. Virus- evolutionary genetic algorithm(VEGA) could be a strategy against this local convergence. Therefore, we use VEGA for fuzzy modeling. In this method, local information is exchanged in population so that population can sustain genetic divergence. finally, to prove the theoretical hypothesis, we provide numerical examples to evaluate the feasibility and generality of fuzzy modeling using VEGA.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.27 no.9
• /
• pp.1463-1468
• /
• 2003

This paper presented the real-time self-tuning learning control based on evolutionary computation, which proves its superiority in finding of the optimal solution at the off-line learning method. The individuals of the populations are reduced in order to learn the evolutionary strategy in real-time, and new method that guarantee the convergence of evolutionary mutations is proposed. It is possible to control the control object slightly varied as time changes. As the state value of the control object is generated, evolutionary strategy is applied each sampling time because the learning process of an estimation, selection, mutation is done in real-time. These algorithms can be applied; the people who do not have knowledge about the technical tuning of dynamic systems could design the controller or problems in which the characteristics of the system dynamics are slightly varied as time changes.

• Aerospace Engineering and Technology
• /
• v.7 no.1
• /
• pp.61-67
• /
• 2008

To design flight control law of an unmanned aerial vehicle, automated control gain determination program was developed. The procedure for determination of control gain was formulated as the control gains were designed from the optimal solutions of the optimization problem. PSO algorithm, which is one of the evolutionary computation method, and SQP algorithm, which is one of the nonlinear programming method, are used as optimization problem solver. Thru this technique, computation time required for finding the optimal solution is decreased to 1/5 of that of PSO algorithm and more accurate optimal solution is obtained.