• 제목/요약/키워드: Iterated function systems

검색결과 15건 처리시간 0.017초

진화방식을 이용한 N명 반복적 죄수 딜레마 게임의 협동연합에 관한 실험적 연구 (Experimental Study on Cooperative Coalition in N-person Iterated Prisoner's Dilemma Game using Evolutionary)

  • 서연규;조성배
    • 한국정보과학회논문지:소프트웨어및응용
    • /
    • 제27권3호
    • /
    • pp.257-265
    • /
    • 2000
  • 자연계의 많은 합리적인 이기적 개체들 사이에는 수많은 선택적 갈등이 존재한다. 반복적 죄수의 딜레마(Iterated Prisoner's Dilemma: IPD)게임은 합리적인 이기적 개체들 사이의 경쟁과 협동에 대한 선택적 갈등을 다루는데, 사회, 경제 및 생물 시스템에서 협동의 진화를 연구하는데 사용되어 왔다. 이제까지는 게임자의 수와 협동의 관계, 기계학습의 일환으로서의 전략학습, 그리고 이득함수가 협동에 미치는 영향 등에 관한 연구가 있었는데, 이 논문에서는 진화방식을 이용하여 이득함수에 따른 협동연합의 크기와 지역화가 NIPD(N-player IPD)게임에서 협동의 진화에 미치는 영향에 대해 밝히고자 한다. 시뮬레이션 결과 협동개체에 대한 이득함수의 기울기가 배반에 대한 이득함수의 기울기보다 급하거나 최소 연합의 크기가 작을수록 협동연합의 정도가 높게 나타나며 상호작용하는 이웃의 크기가 작을수록 협동연합의 정도가 높게 진화됨을 알 수 있었다.

  • PDF

STATIONARY SOLUTIONS FOR ITERATED FUNCTION SYSTEMS CONTROLLED BY STATIONARY PROCESSES

  • Lee, O.;Shin, D.W.
    • 대한수학회지
    • /
    • 제36권4호
    • /
    • pp.737-746
    • /
    • 1999
  • We consider a class of discrete parameter processes on a locally compact Banach space S arising from successive compositions of strictly stationary random maps with state space C(S,S), where C(S,S) is the collection of continuous functions on S into itself. Sufficient conditions for stationary solutions are found. Existence of pth moments and convergence of empirical distributions for trajectories are proved.

  • PDF

MEASURE DERIVATIVE AND ITS APPLICATIONS TO $\sigma$-MULTIFRACTALS

  • Kim, Tae-Sik;Ahn, Tae-Hoon;Kim, Gwang-Il
    • 대한수학회지
    • /
    • 제36권1호
    • /
    • pp.229-241
    • /
    • 1999
  • The fractal space is often associated with natural phenomena with many length scales and the functions defined on this space are usually not differentiable. First we define a $\sigma$-multifractal from $\sigma$-iterated function systems with probability. We introduce the measure derivative through the invariant measure of the $\sigma$-multifractal. We show that the non-differentiable function on the $\sigma$-multifractal can be differentiable with respect to this measure derivative. We apply this result to some examples of ordinary differential equations and diffusion processes on $\sigma$-multifractal spaces.

  • PDF

A Window on the Beauty of Fractal Images: TI-92

  • Kwon, Oh-Nam
    • 한국수학교육학회지시리즈D:수학교육연구
    • /
    • 제5권1호
    • /
    • pp.1-12
    • /
    • 2001
  • Generating fractal images by graphing calculators such as TI_92 combines several important features, which convey the excitement of a living, changing mathematics appropriate to secondary or post-secondary students. The topic of fractal geometry can be illustrated using natural objects such as snowflakes, leaves and ferns. These complex and natural forms are often striking fantastic and beautiful. The examples highlight the fact that complex, natural behaviors can result from simple mathematical rules such as those embodied in iterated function systems(IFS). The visual splendor beauty of fractals, in concert with their ubiquity in nature, revels the intellectual beauty of nonlinear mathematics in a compelling way. The window is now open for students to experience and explore some of the wonder of fractal geometry.

  • PDF

공기역학적 성능 향상을 위한 플랩의 최적 위치 선정 (SELECTION OF THE OPTIMAL POSITION OF THE FLAP FOR THE IMPROVEMENT OF AERODYNAMIC PERFORMANCE)

  • 강형민;박영민;김철완;이창호
    • 한국전산유체공학회지
    • /
    • 제18권4호
    • /
    • pp.41-46
    • /
    • 2013
  • The selection of the optimal position of the flap was performed in order to improve the aerodynamic performance during the take-off and landing processes of aircraft. For this, the existing airfoils of the main wing and flap are selected as the baseline model and the lift coefficients (cl) according to angle of attacks (AOA) were calculated with the change of the position of flap airfoil. The objective function was defined as the consideration of the maximum cl, lift to drag ratio and cl at certain AOA. Then, at 121 experimental points within $20mm{\times}20mm$ domain, two dimensional flow simulations with Spalart-Allmaras turbulence model were performed concerning the AOA from 0 to 15 degree. If the optimal position was located at the domain boundary, the domain moved to the optimal position. These processes were iterated until the position was included in the inside of the domain. From these processes, the flow separation at low AOA was removed and cl increased linearly comparing with that of the baseline model.