Steel sheet incremental cold shaping improvements using hybridized genetic algorithms with support vector machines and neural networks

Abstract

The complexity and difficulties in modelling the most of nowadays real world problems increase as the computational capacity does, specially in those processes where relatively new technology arises. One of such processes is the steel sheet incremental cold shaping. The steel sheet incremental cold shaping process is a new technique for shaping metal sheets when a reduced amount of pieces per lots should be manufactured. As it is a relatively new technique, there is a lack of knowledge in defining the operating conditions, so in order to fit them, before manufacturing a lot a trial and error stage is carried out. A decision support system to reduce the cost of processing and to assist in defining the operating conditions should be studied. This study focus on the analysis and design of the decision support system, and as it is going to be shown, the most suitable features have been found using a wrapper feature selection method, in which genetic algorithms support vector machines and neural networks are hybridized. Some facts concerning the enhanced experimentation needed and the improvements in the algorithm are drawn. © 2011 Springer-Verlag Berlin Heidelberg.