PREDICTION AND EVALUATION OF THE MOLDS STATE USING THE TOOLS OF ARTICICIAL INTELLIGENCE

Abstract

In this paper is presented part of the results from grant project, which is concern the area of primary cooling in continuous steel casting device, especially crystallizer’s surface quality. It was presented part results from crystallizer’s defect catalogue and Methodology of their evaluation. In terms of this paper was presented a method to evaluate narrow crystallizer’s desk qualities, which are dismantled when they are in maintenance department. In this paper is described proposal and development of monitoring and information system for support of crystallizer maintenance control and increase quality of processes which are performing basic functions of monitoring and localization of crystallizer’s desks, crystallizer lifetime prediction and prediction of blank defects. Target of the system is to optimize preventive maintenance of crystallizers desks, with utilizing maximum count of relevant data from technological and maintenance processes. The Department of Automation and Computing in Metallurgy, VSB-TU Ostrava has long dealt with the challenge of life of technical systems. In the frame of this solution it is solved an area of metallurgical objects life. One of these objects is the mold of the continuous steel casting device. The current solution results shows that is necessary to use knowledge, not based only on objective qualitative knowledge, so then knowledge subjective and heuristics. The results are the constructions processes using effective model structures, the same principles which utilize human experts. This approach to solving complex systems leading to deployment of systems based on artificial intelligence methods.