Journal of Korean Society of Steel Construction (한국강구조학회 논문집)

Korean Society of Steel Construction (한국강구조학회)
권호 표지

Prediction of Dynamic Response of Structures Using CMAC

CMAC을 이용한 구조물의 동적응답 예측

  • 김동현 (한국과학기술원 토목공학과) ;
  • 김현택 (한국과학기술원 토목공학과) ;
  • 이인원 (한국과학기술원 토목공학과)
  • Received : 2000.02.18
  • Published : 2000.10.27
Download PDF
⟨ Previous Next ⟩

Abstract

Cerebellar model articulation controller (CMAC) is introduced and used for the identification of structural dynamic model. CMAC has fascinating features in learning speed. It can learn structural response within a few seconds. Therefore it is suitable for the real time identification structures. Real time identification is required in the control of structure which may be damaged or undergo severe change in mechanical properties due to shrinkage or relaxation etc. In numerical examples, it is shown that CMAC trained with the dynamic response of three-story building can predict responses under not trained earthquakes with allowable error. Finally, CMAC has great potential in structural and control engineering.

CMAC을 이용하여 구조물의 지진응답을 예측하였다. CMAC은 매우 빠른 학습성능을 가지고 있는 것이 장점이며 구조물의 동적응답을 학습함에 있어서도 수 초 이내에 만족할 만한 정도로 학습을 끝낸다. 따라서 실시간 학습을 필요로 하는 분야에 매우 효과적으로 사용될 수 있다. 실시간 응답학습은 장기거동 등으로 역학적 특성이 변하거나 손상을 입은 구조물의 적응제어 등이 있다. 수치해석에서는 3층 전단건물의 지진응답을 CMAC을 통하여 학습하였으며 학습은 매우 빠르게 완수 되었다. 결론적으로 CMAC은 구조물의 진동제어 분야에서 매우 효과적으로 사용될 수 있는 인공지능의 하나이다.

Keywords

References

  1. J. Dyn. Sys., Measurement & Control, Trans. ASME v.97 no.3 A new approach to manipulator controller: the cerebellar model articulation controller (CMAC) Albus, J.S.
  2. J. Engrg. Mech., ASCE v.124 no.3 Nonlinear structural control using neural networks Bani-Hani, K.;Ghaboussi, J.
  3. J. Engrg.Mech., ASCE Application of CMAC neural network to real time identification of nonlinear structures Kim, D.H.;Jo, J.S.;Lee, I.W.
  4. Earthquake Engrg. and Struct. Dynamics v.25 Modelling unknown structural systems through the use of neural networks Chassiakos, A.G.;Masri, S.F.
  5. J. Appl. Mech. Trans. ASME v.65 On-line identification of hysteretic systems Chassiakos, A.G.;Masri, S.F.;Smyth, A.W.;Caughey, T.K.
  6. J. Engrg. Mech., ASCE v.121 no.12 Neural network for structural dynamic model identification Chen, H.M.;Qi, G.Z.;Yang, J.C.S.;Amini, F.
  7. J. Computing in Civ. Engrg., ASCE v.9 no.2 Neural network for structural control Chen, H.M.;Tsai, K.H.;Qi, G.Z.;Yang, J.C.S.
  8. Neural Networks v.9 no.7 CMAC with general basis functions Chiang, C.T.;Lin, C.S.
  9. IEEE Trans. Neural Networks v.9 no.6 Generalizing CMAC architecture and training Francisco, J.G.;Anibal, R.F.;Antonio, A.
  10. J. Engrg. Mech., ASCE v.121 no.4 Active control of structures using neural networks Ghaboussi, J.;Joghataie, A.
  11. IEICE Trans. Fundamentals v.E81-A no.12 New high order association memory system based on Newtons forward interpolation Hama, H.;Xing, C.;Liu, Z.
  12. Earthquake Engrg. and Struct. Dynamics v.27 Control of structural seismic response by self-recurrent neural network (SRNN) He, Y.;Wu, J.
  13. J. Computing in Civ. Engrg., ASCE v.13 no.1 MS_CMAC neural network learning model in structural engineering Hung, S.L.;Jan, J.C.
  14. J. Engrg. Mech., ASCE v.126 no.2 An optimal control using neural networks Kim, J.T.;Jung, H.J.;Lee, I.W.
  15. IEEE Trans. Automatic Control. v.39 no.6 Robust control of robot manipulators with parametric uncertainty Koo, K.M.;Kim, J.H.
  16. IEEE Control System Neural network control of automotive fuel-iniection systems Michael, M.;James, S.;Cho, D.I.
  17. J. Engrg. Mech., ASCE v.122 no.10 Actuator dynamics and delay compensation using neurocontrollers Nikzad, K.;Ghaboussi, J.;Paul, S.L.
  18. Active structural control: theory and practice Soong, T.T.
  19. IEEE Control Systems Theory and development of higher-order CMAC neural networks Stephen, H.L.;David, A.H.;Jack, J.G.