Abstract
A general purpose implementation of the Tabu Search metaheuristic, called Universal Tabu Search, is used to optimally design a Locally Recurrent Neural Network architecture. Indeed, the design of a neural network is a tedious and time consuming trial and error operation that leads to structures whose optimality is not guaranteed. In this paper, the problem of choosing the number of hidden neurons and the number of taps and delays in the FIR and IIR network synapses is formalised as an optimisation problem whose cost function to be minimised is the network error calculated on a validation data set. The performances of the proposed approach have been tested on the design problem of a Neural Network controller of a Custom Power protection device.
Similar content being viewed by others
References
Back, D. and Tsoi, A. C.: FIR and IIR synapses, a new neural network architecture for time series modelling, Neural Computation 3 (1991), 375-385.
Back, D. and Tsoi, A. C.: A simplified gradient algorithm for IIR synapse multilayer perceptrons, Neural Computation 5 (1993), 456-462.
Back, D., Wan, E. A., Lawrence, S., and Tsoi, A. C.: A unifying view of some training algorithms for multilayer perceptrons with FIR filter Synapses, Neural Computation 8(1) (1991).
Battiti, R. and Tecchiolli, G.: The reactive tabu search, ORSA J. Computing 6(2) (1994), 126-140.
Campolucci, P., Piazza, F., and Uncini A.: On-line learning algorithms for neural networks with IIR synapses, in: Proc. of IEEE Internat. Conf. on Neural Networks ICNN-95, Perth, December 1995.
Campolucci, P., Piazza, F., and Uncini A.: Causal backpropagation through time for locally recurrent neural networks, in: Proc. of ISCAS-96, IEEE Internat. Symposium on Circuit and Systems, Atlanta, USA, May 1996.
Campolucci, P., Uncini, A., Piazza, F., and Bhaskar, D. R.: On-line learning algorithms for locally recurrent neural networks, IEEE Trans. Neural Networks 10(2) (1999).
Cannas, B., Cincotti, S., Fanni, A., Marchesi, M., Pilo, F., and Usai, M.: Performance analysis of locally recurrent neural networks, COMPEL-Internat. J. Comput. Math. Electrical Electronic Engrg 17(5/6) (1998), 708-716.
Cannas, B., Cincotti, S., Marchesi, M., and Pilo, F.: Chua's circuit can be simulated by locally recurrent neural networks, in: Chaos, Solitons and Fractals, Elsevier Science Publishers, Oxford, UK.
Cannas, B., Celli, G., Marchesi, M., and Pilo, F.: Neural networks for power system condition monitoring and protection, Neurocomputing 23 (1998), 111-123.
Celli, G., Marchesi, M., Mocci, F., and Pilo, F.: Applications of neural networks in power distribution systems diagnosis and control, in: Proc. of the 32nd Universities Power Engineering Conf. UPEC '97, Manchester, UK, 10-12 September 1997, pp. 523-526.
Celli, G., Pilo, F., Sannais, R., and Tosi, M.: Voltage quality improvement by custom power devices: Applications of solid-state breakers and neural controllers, in: Proc. of SPEEDAM '98 Conf., Sorrento, Italy, 3-5 June 1998.
Celli, G. and Pilo, F.: A custom power protection device controlled by a neural network relay, in: Proc. of IEEE-PES Summer Meeting, Seatle, USA, July 2000, pp. 1384-1389.
Douglas, J.: Custom power: optimizing distribution services, The EPRI J. 21(3) (1996), 6-15.
Dugan, R. C., McGranaghan, M. F., and Beaty, H. W.: Electric Power Systems Quality, McGraw-Hill, New York, 1996.
Fanni, A., Bibbo', S., Giua, A., and Matta, A.: A general purpose Tabu Search code: An application to digital filters design, in: IEEE Internat. Conf. on Systems, Man, and Cybernetics, San Diego, CA, October 11-14, 1998.
Glover, F. and Laguna, M.: Tabu Search Modern Heuristic Techniques for Combinatorial Problems, Blackwell Scientific Publications, Oxford, 1993, pp. 70-150.
Glover, F.: Tabu search fundamentals and uses, Technical Report, University of Colorado, Boulder, 1994.
Glover, F.: Tabu search-part I, ORSA J. Computing 1(3) (1989), 190-206.
Glover, F.: Tabu search-part II, ORSA J. Computing 2(1) (1990), 4-32.
Glover, F. and Laguna M.: Tabu Search, Kluwer Academic, Dordrecht, 1997.
Goldberg, D. E.: Genetic Algorithms in Search, Optimization and Machine Learning, Addison-Wesley, 1989.
Hingorani, N. G.: Introducing custom power, IEEE Spectrum (June 1995), 41-48.
Italian Standard CEI 11-20, No. 1444, January 1991.
Khotanzad, Afkhami-Rohani, R., Lu, T., Abaye, A., Davis M., and Maratukulam D. J.: ANNSTLF-A neural-network-based electric load forecasting system, Trans. Neural Networks 8(4) (1997).
Kirkpatrick, S., Gelatt, C. D., Jr., and Vecchi, M. P.: Optimization by simulated annealing, Science 220 (May 1983), 671-680.
Pittorino, L. and Enslin, J.: Evaluation of converter topologies and controllers for power quality solutions, in: Proc. of the 1st Southern African Power Quality Conf., June 1997.
Smith, M.: Neural Networks for Statistical Modeling, Van Nostrand Reinhold, 1993.
Wan, E. A.: Finite impulse response neural networks with applications in time series prediction, Ph.d Dissertation, Stanford University, 1996.
Wan, E. A.: (1990): Temporal backpropagation for FIR neural networks, in: Proc. of Internat. Joint Conf. on Neural Networks, San Diego, June 1990, pp. 575-580.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Cannas, B., Celli, G., Fanni, A. et al. Automated Recurrent Neural Network Design of a Neural Controller in a Custom Power Device. Journal of Intelligent and Robotic Systems 31, 229–251 (2001). https://doi.org/10.1023/A:1012002911516
Issue Date:
DOI: https://doi.org/10.1023/A:1012002911516