# MERGING OF ADVICES FROM MULTIPLE ADVISORY SYSTEMS - With Evaluation on Rolling Mill Data

### Pavel Ettler, Josef Andrýsek, Václav Šmídl, Miroslav Kárný

#### Abstract

The problem of evaluation of advisory system quality is studied. Specifically, 18 advisory strategies for operators of a cold rolling mill were designed using different modelling assumptions. Since some assumptions may be more appropriate in different working regimes, we also design a new advising strategy based on the on-line merging of advices. In order to measure actual suitability of the advisory systems, we define two measures: operator’s performance index and coincidence of the observed operator’s actions with the advices. A time-variant model of advisory system suitability is proposed. Merging of the advices is achieved using Bayesian theory of decision-making. Final assessment of the original advisory systems and the new system is performed on data recorded during 6 months of operation of a real rolling mill. This task is complicated by the fact that the operator did not follow any of the recommendations generated by the advisory systems. Validation was thus performed with respect to the proposed measures. It was found that merging of the advising strategies can significantly improve quality of advising. The approach is general enough to be used in many similar problems.

#### References

- Ettler, P. and Jirkovský, F. (1991). Digital controllers for škoda rolling mills. In Warwick, K., Kárný, M., and Halousková, A., editors, Lecture Notes: Advanced Methods in Adaptive Control for Industrial Application (Joint UK-CS seminar) , volume 158, pages 31- 35. Springer Verlag.
- Ettler, P., Kárný, M., and Guy, T. V. (2005a). Bayes for rolling mills: From parameter estimation to decision support. In 16th IFAC World Congress, Praha, CZ.
- Ettler, P., Kárný, M., and Guy, T. V. (2005b). Bayes for rolling mills: From parameter estimation to decision support. In Accepted for the 16th IFAC World Congress, Praha, CZ.
- Kárný, M., Böhm, J., Guy, T., Jirsa, L., Nagy, I., Nedoma, P., and Tesar?, L. (2005). Optimized Bayesian Dynamic Advising: Theory and Algorithms. Springer, London. to appear.
- Kárný, M., Kracík, J., and Guy, T. (2007). Cooperative decision making without facilitator. In Andrievsky B.R., F. A., editor, IFAC Workshop "Adaptation and Learning in Control and Signal Processing" /9./. IFAC.
- Kulhavý, R. and Zarrop, M. B. (1993). On a general concept of forgetting. International Journal of Control, 58(4):905-924.
- Maz'ya, V. and Schmidt, G. (2001). On approximate approximations using Gaussian kernels. IMA Journal of Numerical Analysis, 16(1):13-29.
- Quinn, A., Ettler, P., Jirsa, L., Nagy, I., and Nedoma, P. (2003). Probabilistic advisory systems for dataintensive applications. International Journal of Adaptive Control and Signal Processing, 17(2):133-148.
- Raftery, A., Madigan, D., and J.A.Hoeting (1997). Bayesian model averaging for linear regression models. Journal of Thhe American Statistical Association, 97(437):179-191.
- Titterington, D., Smith, A., and Makov, U. (1985). Statistical Analysis of Finite Mixtures. John Wiley, New York.
- Nedoma, P., Kárný, M., Böhm, J., and Guy, T. V. (2005). Mixtools Interactive User's Guide. Technical Report 2143, ÚTIA AV C?R, Praha.

#### Paper Citation

#### in Harvard Style

Ettler P., Andrýsek J., Šmídl V. and Kárný M. (2008). **MERGING OF ADVICES FROM MULTIPLE ADVISORY SYSTEMS - With Evaluation on Rolling Mill Data** . In *Proceedings of the Fifth International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO,* ISBN 978-989-8111-30-2, pages 66-71. DOI: 10.5220/0001500900660071

#### in Bibtex Style

@conference{icinco08,

author={Pavel Ettler and Josef Andrýsek and Václav Šmídl and Miroslav Kárný},

title={MERGING OF ADVICES FROM MULTIPLE ADVISORY SYSTEMS - With Evaluation on Rolling Mill Data},

booktitle={Proceedings of the Fifth International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO,},

year={2008},

pages={66-71},

publisher={SciTePress},

organization={INSTICC},

doi={10.5220/0001500900660071},

isbn={978-989-8111-30-2},

}

#### in EndNote Style

TY - CONF

JO - Proceedings of the Fifth International Conference on Informatics in Control, Automation and Robotics - Volume 1: ICINCO,

TI - MERGING OF ADVICES FROM MULTIPLE ADVISORY SYSTEMS - With Evaluation on Rolling Mill Data

SN - 978-989-8111-30-2

AU - Ettler P.

AU - Andrýsek J.

AU - Šmídl V.

AU - Kárný M.

PY - 2008

SP - 66

EP - 71

DO - 10.5220/0001500900660071