An Evolutionary Optimization Approach for Bulk Material Blending Systems

Michael P. Cipold^{1, 2}, Pradyumn Kumar Shukla¹, Claus C. Bachmann², Kaibin Bao¹, and Hartmut Schmeck¹

¹Institute AIFB, Karlsruhe Institute of Technology, Karlsruhe, D-76128, Germany

²J&C Bachmann GmbH, Bad Wildbad, D-75323, Germany

Abstract. Bulk material blending systems still mostly implement static and non-reactive material blending methods like the well-known Chevron stacking. The optimization potential in the existing systems which can be made available using quality analyzing methods as online X-ray fluorescence measurement is inspected in detail in this paper using a multi-objective optimization approach based on steady state evolutionary algorithms. We propose various Baldwinian and Lamarckian repair algorithms, test them on real world problem data and deliver optimized solutions which outperform the standard techniques.

Keywords: Bulk Material Blending, Multi-objective Evolutionary Algorithms, Chevron Stacking

LNCS 7491, p. 478 ff.

Full article in PDF | BibTeX

lncs@springer.com
© Springer-Verlag Berlin Heidelberg 2012