《钢材表面质量检查系统技术交流资料6》.pdfVIP

The application of hyperspectral image processing to the steel foundry process. Authors: José A. Gutiérrez Olabarria. Engineer, Tecnalia- Infotech josea@robotiker.es Artzai Picón Ruiz. Ph.D. in Industrial Engineering, Tecnalia- Infotech, apicon@robotiker.es Sergio Rodriguez Vaamonde, Engineer, Tecnalia-Infotech srodriguez@robotiker.es Robotiker- Tecnalia Parque Tecnológico Edif. 202 48170 Zamudio (Biscay) Spain Ignacio Girbau, Engineer, Arcelor – Mittal Olaberria ignacio.girbau@ Arcelor – Mittal Olaberria Ctra. Madrid-Irún, KM. 417. 20212 Olaberria (Gipuzkoa) Spain 1. Summary In the steel industry many technologies are applied in order to optimise the cost of the production process and to improve the quality of the final product. Despite these efforts, there are still areas where the parameters of the process are established by subjective evaluations. More specifically, this is the case in the slag characterisation processes in the Electrical Arc Furnace (EAF) and in the Ladle Furnace (LF). In the EAF process, to be precise, the chemical composition of slag is not known during the process as the samples must be analysed in special spectrometers and this requires a previous treatment and preparation. In this process, the knowledge of chemical composition of slag would allow to know the basicity and oxidation of the slag. At present, both of these characteristics are only known after laboratory tests. During the LF process, the estimation of the degree of deoxidation and desulfuration of steel is based on the visual evaluation (Figure 1) of the slag samples and on the chemical comp