Neural Networks and Particle Swarm Optimization for Function Approximation in Tri-SWACH Hull Design

ABSTRACT

Tri-SWACH is a novel multihull ship design that is well suited to a wide range of industrial, commercial, and military applications, but which because of its novelty has few experimental studies on which to base further development work. Using a new form of particle swarm optimization that incorporates a strong element of stochastic search, Breeding PSO, it is shown it is possible to use multilayer nets to predict resistance functions for Tri-SWACH hullforms, including one function, the Residual Resistance Coefficient, which was found intractable with previously explored neural network training methods.

References

1. Gougouldis, G. 2008. The utilization of artificial neural networks in marine applications: an overview. Naval Engineering Journal 120, 3 (Feb. 2008), 19--26.
2. Couser, P., Mason, A., Mason, G., Smith, C, and Von Konsky, B. 2004. Artificial neural networks for hull resistance prediction. In Proc. 3rd International Conference on Computer Applications in the Marine Industries (Siguënza, Spain, May 09-12, 2004). COMPIT '04. Springer, Netherlands, 391--402.
3. Royce, R., Mouravieff, A., and Zuzick. A. 2011. A trimaran resistance artificial neural network. In Proceedings of the 11th International Conference on Fast Ship Transportation (Honolulu, Hawaii, USA, Sept. 26--29, 2011). FAST '11. American Soc. Naval Engineers, Alexandria, VA, 717--726.
4. Carter, A., Muk-Pavic, N., and McDonald, T. 2013. Resistance prediction using artificial neural networks for preliminary Tri-SWACH design. Trans. RINA, Vol. 156, Part A:, Intl. J. Maritime Engineering No.263, 141--152.
5. Chaggar, N. 2014. Tri-SWACH Resistance Prediction Using Artificial Neural Networks. MSc thesis, Dept. of Mechanical Engineering, University College London.
6. Klag, J., and McMahon, I. 2007. Calm Water Resistance Study of a Novel Trimaran. Doctoral thesis. Webb Institute, Glen Cove, USA.
7. Kennedy, J., and Eberhart, R. 1995. Particle swarm optimization. In IEEE International Conference Symposium on Neural Networks (Perth, Australia, Nov. 27-Dec. 01, 1995). IEEE Press, New York, NY, 1942--1948.
8. Ilonen, J., Kamarainen, J.-K., and Lampinen, J. 2006. Differential evolution training algorithm for feed-forward neural networks. Neural Processing Letters 17, 1 (June 2006), 93--105.
9. Storn, R. and Price, K. 1997. Differential evolution -- a simple and efficient heuristic for global optimization over continuous spaces. Journal of Global Optimization 11, 4 (Dec. 1997), 341--359.
10. Liang, J.J., Qin, A.K., and Baskar, S. 2006. Comprehensive learning particle swarm optimizer for global optimization of multimodal functions. IEEE Trans. Evolutionary Computation 10, 3 (June 2006), 281--295.
11. Liang, J.J., Kai Qin, A., Suganthan, P.N., and Baskar, S. 2004. Evaluation of comprehensive learning particle swarm optimizer. In Proc. 11th International Conference on Neural Information Processing (Calcutta, India, Nov. 22--25, 2004). ICONIP 2004. Springer-Verlag, Berlin Heidelberg, Germany, 230--235.
12. Das, S., Abraham, A., Chakraborty, U., K., and Konar, A. 2009. Differential evolution using a neighborhood-based mutation operator. IEEE Trans. Evolutionary Computation 13, 3 (June 2009), 526--553
13. Bengio, Y. 2009. Learning Deep Architectures for AI (Foundations and Trends in Machine Learning). Now Publishers Inc., Hanover, MA.
14. Ciresan, D., Meier, U., Masci, J., and Schmidhuber, J. 2012. Multi-column deep neural network for traffic sign classification. Neural Networks, 32 (Feb. 2009), 333--338.
15. Fortier, N., Sheppard, J.W., and Pillai, K.G. 2012. DOSI: Training artificial neural networks using overlapping swarm intelligence with local credit assignment. In Proc. 13th International Symposium on Advanced Intelligent Systems (Kobe, Japan, Nov. 20-24, 2012). ISIS '12. IEEE Press, New York, NY, 1420--1425.
16. Piotrowski, A.P. and Napiorkowski, J.J. 2011. Optimizing neural networks for river flow forecasting -- evolutionary computation methods versus the Levenberg--Marquardt approach. J. Hydrology, 407, 1 (Sept. 2011), 12--27.

Index Terms

• Computing methodologies

  • Machine learning

    • Machine learning approaches

      • Neural networks

• Hardware

  • Emerging technologies

    • Circuit substrates

      • Cellular neural networks

• Mathematics of computing

  • Mathematical analysis

    • Functional analysis

      • Approximation

• Theory of computation

  • Design and analysis of algorithms

    • Approximation algorithms analysis

  • Randomness, geometry and discrete structures

    • Computational geometry

  • Theory and algorithms for application domains

    • Machine learning theory