• 한국지능시스템학회논문지
• /
• 제8권3호
• /
• pp.1-8
• /
• 1998

본 논문에서는 유전자 프로그래밍의 성능을 향상시키기 위하여 강화학습법에 기반한 강화 유전자 프로그래밍을 제안한다. 트리구조와 프로그램을 염색체로 가지는 유전자 프로그래밍(GP)은 다른 진화 알고리즘에 비해 염색체의 크기에 제한이 없기 때문에 표현력에 융통성이 많다는 장점이 있다. 그러나 이러한 특징은 반대고 교차 및 돌연변이 연산에 있어서 수렴성을 떨어뜨리는 단점을 나타낸다. 따라서 유전자 프로그래밍은 다른 진화알고리즘에 비해 개체군의 크기 및 진화 세대수를 크게 잡는 것이 일반적이다. 본 논문에서는 유전자 프로그래밍의 이러한 성질을 개선하기 위해서 프로그램에 강화신호를 주어 이것의 보답/벌칙의 정도에 기반한 교차 및 돌연번이 연산을 실행하는 방법을 제안한다. 제안된 방법은 인공개미(Artificial Ant)문제에 적용하여 그 유효성을 입증한다.

• 한국CDE학회논문집
• /
• 제19권1호
• /
• pp.1-11
• /
• 2014

In the initial design stage, the technology for estimating and managing the weight of a floating offshore structure, such as a FPSO (Floating, Production, Storage, and Off-loading unit) and an offshore wind turbine, has a close relationship with the basic performance and the price of the structure. In this study, using the genetic programming (GP), being used a lot in the approximate estimating model and etc., the weight estimation model of the floating offshore structure was studied. For this purpose, various data for estimating the weight of the floating offshore structure were collected through the literature survey, and then the genetic programming method for developing the weight estimation model was studied and implemented. Finally, to examine the applicability of the developed model, it was applied to examples of the weight estimation of a FPSO topsides and an offshore wind turbine. As a result, it was shown that the developed model can be applied the weight estimation process of the floating offshore structure at the early design stage.

• 한국지능시스템학회논문지
• /
• 제20권5호
• /
• pp.603-609
• /
• 2010

본 논문은 SURF(Speeded Up Robust Features)를 기반으로 한 대상 물체 인식 알고리즘과 GP(Genetic Programming)를 기반으로 한 직진, 회전, 정지, 후진 걸음새(gait) 자동 생성을 각각 구현한다. 그리고 이를 결합 하여, 대상을 인식하고 자율적으로 접근 및 추종할 수 있는 인식 기반 지능적인 보행 기법을 제안한다. 4족 보행 로봇의 걸음새는 GP를 사용하여 각 관절의 궤적에 대한 회귀분석으로 생성한다. 고속의 특징점 검출에 적합한 SURF를 사용해서 물체의 위치와 크기를 인식하고, 물체까지의 거리를 계산한다. 4족 보행로봇의 물체 인식 및 이를 통한 자율접근 보행 실험은 ODE(Open Dynamics Engine) 기반의 Webots 시뮬레이션과 실제 로봇에 대해서 수행된다.

• 한국지능시스템학회논문지
• /
• 제16권6호
• /
• pp.677-682
• /
• 2006

멀티-도메인 공학시스템은 전기, 기계, 유압, 열등의 구성요소를 포함하고, 시스템 구성이 복잡하여 설계에 많은 어려움을 가지고 있다. 최적의 설계를 위해서는 각 도메인에 대한 통합된 설계 방법과 자동적이고 효율적인 탐색방법이 요구된다. 본 논문은 도메인에 독립적인 모델링 도구인 본드 그래프(Bond Graph)와 대규모 공간 해의 탐색에 접합한 진화 알고리즘의 일종인 유전 프로그래밍(Genetic Programming)를 결합하여 멀티 도메인 동적시스템에 대한 디자인 해를 자동적으로 생성해주는 설계 방법을 제시하였다. 제안된 설계방법의 효용성을 입증하기 위해서 3가지 서로 다른 도메인을 가진 아나로그 필터, 전동프린터 드라이브, 에어펌프 시스템에 대한 설계 결과가 기술된다.

• 예술인문사회 융합 멀티미디어 논문지
• /
• 제8권4호
• /
• pp.259-267
• /
• 2018

부분집합 합 문제는 유한개의 정수로 이루어진 집합이 있을 때 이 집합의 부분집합 중에서 그 집합의 원소들의 합이 특정 값이 되는 경우가 있는지를 알아내는 문제로, 잘 알려진 다항식 시간 내에 풀기 어려운 NP-완비 문제이다. 유전 알고리즘은 선택과 교차, 돌연변이 등의 연산을 통해 주어진 문제의 최적해를 구하는 알고리즘이다. 동적 계획법은 주어진 문제를 풀기 위해서 문제를 하나 또는 여러 개의 하위 문제로 나누어 풀이하는 방법이다. 본 논문에서는 부분집합 합 문제를 풀이하는 유전 알고리즘을 설계 및 구현하고, 답을 찾는 데까지 걸리는 시간 성능을 동적 계획법의 경우와 실험적으로 비교하였다. 양의 정수인 원소 63 개를 가진 집합에서 '쉬움'과 '어려움'의 난이도를 고려하여 총 17 개의 문제를 선정하고, 이 문제들을 풀이하는 두 알고리즘의 성능을 비교하는 실험을 진행하였다. 17 개의 문제 중 13 개의 문제에서 본 논문에서 제시한 유전 알고리즘은 동적 계획법과 비교하여 약 84%가 우수한 시간 성능을 보였다.

• Journal of Information Processing Systems
• /
• 제14권2호
• /
• pp.487-498
• /
• 2018

The purpose of the high-speed railway construction is to better satisfy passenger travel demands. Accordingly, the design of the train working plan must also take a full account of the interests of passengers. Aiming at problems, such as the complex transport organization and different speed trains coexisting, combined with the existing research on the train working plan optimization model, the multiobjective bi-level programming model of the high-speed railway passenger train working plan was established. This model considers the interests of passengers as the center and also takes into account the interests of railway transport enterprises. Specifically, passenger travel cost and travel time minimizations are both considered as the objectives of upper-level programming, whereas railway enterprise profit maximization is regarded as the objective of the lower-level programming. The model solution algorithm based on genetic algorithm was proposed. Through an example analysis, the feasibility and rationality of the model and algorithm were proved.

• 한국정밀공학회지
• /
• 제19권10호
• /
• pp.66-76
• /
• 2002

Automated welding and soldering are an important manufacturing issue in order to lower the cost, increase the quality, and avoid labor problems. An off-line programming, OLP, is one of the powerful methods to solve this kind of diversity problem. Unless an OLP system is ready for the path optimization in welding and soldering, the waste of time and cost is unavoidable due to inefficient paths in welding and soldering processes. Therefore, this study attempts to obtain path optimization using a genetic algorithm based on artificial intelligences. The problem of welding path optimization is defined as a conventional TSP (traveling salesman problem), but still paths have to go through welding lines. An improved genetic algorithm was suggested and the problem was formulated as a TSP problem considering the both end points of each welding line read from database files, and then the transit problem of welding line was solved using the improved suggested genetic algorithm.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제13권4호
• /
• pp.1765-1794
• /
• 2019

Genetic Programming (GP) is an intelligence technique whereby computer programs are encoded as a set of genes which are evolved utilizing a Genetic Algorithm (GA). In other words, the GP employs novel optimization techniques to modify computer programs; imitating the way humans develop programs by progressively re-writing them for solving problems automatically. Trial programs are frequently altered in the search for obtaining superior solutions due to the base is GA. These are evolutionary search techniques inspired by biological evolution such as mutation, reproduction, natural selection, recombination, and survival of the fittest. The power of GAs is being represented by an advancing range of applications; vector processing, quantum computing, VLSI circuit layout, and so on. But one of the most significant uses of GAs is the automatic generation of programs. Technically, the GP solves problems automatically without having to tell the computer specifically how to process it. To meet this requirement, the GP utilizes GAs to a "population" of trial programs, traditionally encoded in memory as tree-structures. Trial programs are estimated using a "fitness function" and the suited solutions picked for re-evaluation and modification such that this sequence is replicated until a "correct" program is generated. GP has represented its power by modifying a simple program for categorizing news stories, executing optical character recognition, medical signal filters, and for target identification, etc. This paper reviews existing literature regarding the GPs and their applications in different scientific fields and aims to provide an easy understanding of various types of GPs for beginners.

• Computers and Concrete
• /
• 제14권4호
• /
• pp.479-501
• /
• 2014

This paper presents the improvement of the EC-2 and EHE-08 shear strength formulations for concrete beams with shear reinforcement. The employed method is based on the genetic programming (GP) technique, which is configured to generate symbolic regression from a set of experimental data by considering the interactions among precision, accuracy, safety and simplicity. The size effect and the influence of the amount of shear reinforcement are examined. To develop and verify the models, 257 experimental tests on concrete beams from the literature are used. Three expressions of considerable simplicity, which significantly improve the shear strength prediction with respect to the formulations of the different studied codes, are proposed.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 1998년도 The Third Asian Fuzzy Systems Symposium
• /
• pp.170-175
• /
• 1998

Transporting a large table using multiple robotic agents requires at least two group behaviors of homing and herding which are to bo coordinated in a proper sequence. Existing GP methods for multi-agent learning are not practical enough to find an optimal solution in this domain. To evolve this kind of complex cooperative behavior we use a novel method called fitness switching. This method maintains a pool of basis fitness functions each of which corresponds to a primitive group behavior. The basis functions are then progressively combined into more complex fitness functions to co-evolve more complex behavior. The performance of the presented method is compared with that of two conventional methods. Experimental results show that coevolutionary fitness switching provides an effective mechanism for evolving complex emergent behavior which may not be solved by simple genetic programming.