Journal of the Institute of Electronics Engineers of Korea CI (전자공학회논문지CI)

The Institute of Electronics and Information Engineers (대한전자공학회)
권호 표지

A Design of the Recurrent NN Controller for Autonomous Mobil Robot by Coadaptation of Evolution and Learning

진화와 학습의 상호 적응에 의한 자발적 주행 로봇을 위한 재귀 신경망 제어기 설계

  • Kim, Dae-Jin (Pohang University of Science and Technology) ;
  • Gang, Dae-Seong (Dept.of Electric Electronics Computer Engineering, Donga University)
  • 김대진 (포항공과대학교 컴퓨터공학과) ;
  • 강대성 (동아대학교 전기전자컴퓨터공학부)
  • Published : 2000.05.01
Download PDF
⟨ Previous Next ⟩

Abstract

This paper proposes how the recurrent neural network controller for a Khepera mobile robot with an obstacle avoiding ability can be determined by co-adaptation of the evolution and learning, The proposed co-adaptation scheme consists of two folds： a population of NN controllers are evolved by the genetic algorithm so that the degree of obstacle avoidance might be reduced through the global searching and each NN controller is trained by CRBP learning so that the running behavior is adapted to its outer environment through the local searching. Experimental results shows that the NN controller coadapted by evolution and learning outperforms its non-learning equivalent evolved by only genetic algorithm in both the ability of obstacle avoidance and the convergence speed reaching to the required running behavior.

본 논문은 장애물 회피 능력을 갖는 자발적 주행 로봇 (Khepera)을 제어하는 재귀 신경망을 진화와 학습의 상호 적응에 의해 결정하는 방안을 제시한다. 제안한 동시 적응 방안은 다음 두 가지 성질을 갖는다. 유전자 알고리즘에 의해 해집단내 여러 개의 신경망 제어기들은 전역적 탐색을 수행하여 점진적으로 장애물과의 충돌이 적게 일어나도록 진화되고, 동시에 각 신경망 제어기는 상보적 재강화 역전파 (CRBP： Complementary Reinforcement Backpropagation) 학습에 의해 국부적 탐색을 수행하여 주행 특성이 로봇이 처한. 외부 환경에 적응되어진다. 실험 결과, 학습과 결합한 진화에 의해 얻어진 신경망 제어기가 진화자체만에 의해 얻어진 신경망 제어기보다 더 나은 충돌 회피 능력을 보여 주며, 원하는 주행 성능에 보다 빨리 도달하는 것을 확인할 수 있다.

Keywords

References

  1. L. Steels, 'Building agents out of autonomous behavior systems,' Building Simulated embodied Agents, New Haven: Lowrence Erlbaum, 1993
  2. L. Meeden, 'An incremental approach to developing intelligent nerual network controllers for robots,' IEEE Transactions on Systems, Man, and Cybernetics,' vol. 26, no. 3, pp. 474-485, 1996 https://doi.org/10.1109/3477.499797
  3. I. Harvey, P. Husbands, and D. Cliff, 'Issues in evolutionary robotics,' From Animals to Animats 2, J.-A. Meyer, H. Roitblat, and S. Wilson, Eds. Cambridge, MA: MlTPress, 1993, pp. 364-373
  4. D. Floreano and F. Mondada, 'Evolution of homing navigation in a real mobile robot,' IEEE Transaction on Systems, Man, and Cybernetics,' vol. 26, no. 3, pp. 396-407, 1996 https://doi.org/10.1109/3477.499791
  5. J. Millan, 'Rapid, safe, and incremental learning of navigation stragegies,' IEEE Transaction on Systems, Man, and Cybernetics,' vol. 26, no. 3, pp. 408-420, 1996 https://doi.org/10.1109/3477.499792
  6. R. K. Belew, J. McInerney, and N. N. Schraudolph, 'Evolving networks : Using the genetic algorithm with connectionist learning,' Artificial Life II, edited by C. G. Langton, J. D. Farmer, S. Rasmussen, and C. E. Taylor, pp. 511-548, Addison-Wesley, 1991
  7. G. F. Miller, P. M. Todd, and S. U. Hedge, 'Designing neural networks using genetic algorithms,' Proceedings of the Third International Conference on Genetic Algorithms, Morgan Kaufmann, 1989
  8. J. B. Lamarck, 'Of the influence of the environment on the activities and habits of animals,' it Zoological Philosophy, Macmillan, London, 1914
  9. D. H. Ackley, and M. L. Littman, 'A case for Lamarckian evolution,' Artificial Life III, edited by C. G. Langton, pp. 3-10, Addison-Wesley, 1994
  10. J. M. Baldwin, 'A new factor in evolution,' The American Naturalist, vol 30, June, 1896
  11. G. E. Hinton and S. J. Nowlan, 'How learning can guide evolution,' Adaptive Individuals in Evolving Populations : Models and Algorithms, edited by R. K. Belew and M. Mitchell, pp. 447-454, Addison Wesley, 1996
  12. D. Parisi, S. Nolfi, 'The influence of learning on evolution,' Adaptive Individuals in Evolving Populations : Models and Algorithms, edited by R. K. Belew and M. Mitchell, pp. 419-430, Addison Wesley, 1996
  13. D. H. Ackley, and M. L. Littman, 'Generalization and scaling in reinforcement learning,' in Advances in Neural Information Processing System 2, D. S. Touretsky, Ed. San Mateo, CA: Morgan Kaufmann, 1990. pp. 550-557
  14. R. Williams, and David Zipser, 'A learning algorithm for continually running fully recurrent neural networks,' Neural Computation, vol. 1, pp. 270-280, 1989
  15. J. H. Holland, Adaptation in natural and artificial systems, University of Michigan Press, 1975
  16. David E. Goldberg, Genetic Algorithms in Search, Optimization and Machine Learning, Addison Wesley Press, 1989
  17. A. H. Wright, 'Genetic algorithms for real parameter optimization,' Foundations of Genetic Algorithms, edited by G. Rawlins, pp. 205-220, Morgan Kaufmann Publishers, 1991
  18. Daijin Kim and Chulhyun Kim, 'Forecasting Time Series with Genetic Fuzzy Predictor Ensemble', IEEE Trans. on Fuzzy Systems, vol. 5, no. 4, 1997 https://doi.org/10.1109/91.649903
  19. Daijin Kim, Sunha Ahn and Dae-Seong Kang, 'Co-adaptation of Self-Organizing Maps by Evolution and Learning,' NeuroComputing, vol. 30, pp. 249-272, 2000 https://doi.org/10.1016/S0925-2312(99)00129-0
  20. F. Mondada, E. Pranzi, and P. Ienne, 'Mobil robot miniaturization: A tool for investigation in control algorithms,' Proceedings of the Thrid International Symopsium on Experimental Robotics, Kyoto, Japan, 1993
  21. O. Michel, 'Khepera Simulator v2.0', User's Manual, University of Nice, 1996
  22. F. Mondada, and D. Floreano, 'Evolution of neural control structures: some experiments on mobile robots,' Robotics and Autonomous Systems, vol 16, pp. 183-195, 1995 https://doi.org/10.1016/0921-8890(96)81008-6