Abstract
An evolutionary algorithm for the development of neural networks with arbitrary connectivity is presented. The algorithm is not based on genetic algorithms, but is inspired by a biological theory of coevolving species. It sets no constraints on the number of neurons and the architecture of a network, and develops network topology and parameters like weights and bias terms simultaneously. Designed for generating neuromodules acting in embedded systems like autonomous agents, it can be used also for the evolution of neural networks solving nonlinear control problems. Here we report on a first test, where the algorithm is applied to a standard control problem: the balancing of an inverted pendulum.
This is a preview of subscription content, log in via an institution.
Preview
Unable to display preview. Download preview PDF.
References
Albrecht, R. F., Reeves, C. R., and Steele, N. C. (eds.), Artificial Neural Nets and Genetic Algorithms, Proceedings of the International Conference in Innsbruck, Austria, 1993, Springer, Wien 1993.
Anderson, C. W. (1989). Learning to control an inverted pendulum using neural networks. IEEE Control Systems Magazine, 9, 31–37.
Anderson, C. W. and Miller W. T. (1990). Challinging Control Problems. In W. T. Miller, R. S. Sutton, and P. J. Werbos, Neural Networks for Control, MIT Press.
Barto, A. G., Sutton, R. S., and Anderson, C. W. (1983). Neuronlike adaptive elements that solve difficult learning control problems. IEEE Transactions on Systems, Man, Cybernetics, 13, 834–846.
Branke, J. (1995). Evolutionary algorithms for neural network design and training. In Proceedings 1st Nordic Workshop on Genetic Algorithms and its Applications, Vaasa, Finland.
Bapi, R. S., D'Cruz, B., and Bugmann, G. (1996) Neuro-resistive grid approach to trainable controllers: A pole balancing example, submitted to Neural Computing and Applications.
Dasgupta, D., and McGregor, D. R. (1993). Evolving neurocontrollers for pole balancing. In S. Gielen and B. Kappen (eds.), ICANN'93 Proceedings of the International Conference on Artificial Neural Networks, Amsterdam 13.–16. Sept. 1993. Berlin: Springer-Verlag, 1993, pp. 834–837.
Dieckmann, U. (1995), Coevolution as an autonomous learning strategy for neuromodules, in: Herrmann, H., Pöppel, E., and Wolf, D. (eds.), Supcrcomputing in Brain Research — From Tomography to Neural Networks, Singapore: World Scientific, (pp.331–347).
Geva, S., and Sitte, J. (1993), A cartpole experiment benchmark for trainable controllers, IEEE Control Systems Magazin, 13, 40–51.
Schaffer, J. D., Whitley, D., and Eshelman, L. J. (1992). Combination of genetic algorithms and neural networks: A survey of the state of the art. In: Proccedings International Workshop on combinations of genetic algorithms and neural networks (COGANN-92), Los Alamitos, CA, IEEE Computer Society Press.
Selfridge, O. G., Sutton, R. S., and Barto, A. G. (1985). Training and tracking in robotics. In Proccedings International Joint Conference on Artificial Intelligence (IJCAI-85), Los Angeles, CA, pp: 670–672.
Widrow, B. (1987), The original adaptive neural net broom-balancer. Proc. IEEE Intern. Symp. Circuits and Systems, 351–357.
Wieland, A. P. (1991). Evolving neural network controllers for unstable systems. In: International Joint Conference on Neural Networks, Seattle, USA, Julyl991. Proccedings Vol.2. Seattle: IEEE Service Center, 1991.
Yao, X. (1993). A review of evolutionary artificial neural networks. International Journal of Intelligent Systems, 8, 539–567.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1997 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Pasemann, F., Dieckmann, U. (1997). Evolved neurocontrollers for pole-balancing. In: Mira, J., Moreno-Díaz, R., Cabestany, J. (eds) Biological and Artificial Computation: From Neuroscience to Technology. IWANN 1997. Lecture Notes in Computer Science, vol 1240. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0032588
Download citation
DOI: https://doi.org/10.1007/BFb0032588
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-63047-0
Online ISBN: 978-3-540-69074-0
eBook Packages: Springer Book Archive