Artificial intelligence to optimize melting processes: an approach combining data acquisition and modeling

Abstract

Melting and recycling of Al alloys involve large amounts of energy and CO2 release. In order to minimize energy consumption and environment impact, a novel approach has been developed and tested for this industrial sector, but it can be extended to other processes and materials. The approach is based on on-line data acquisition and efficient numerical modeling of heat exchanges within a melting furnace. The fast and efficient numerical model, which includes the physical mechanisms of combustion, radiation, conduction and convection, has a few adjustable parameters which are calibrated on-line by a few data acquisition values. A friendly user-interface allows furnace operators to monitor the melting process and optimize mass loading, door opening, heating sequences, etc. The main features of this tool are presented.