BEGIN:VCALENDAR
VERSION:2.0
METHOD:PUBLISH
BEGIN:VEVENT
ORGANIZER;CN=ESTAD 2023:mailto:info@metec-estad.com
LOCATION:Room 19b
SUMMARY:Analysis of vibrations from metallurgical processes to improve process control utilising advanced mathematical algorithms as machine learning
DESCRIPTION:Measurement of the vibrations from metallurgical processes may give an indirect description of the process performance, thereby providing means to improve the overall process control. This approach is applied to the gas stirring of a ladle during vacuum treatment and the AOD converter in the present work. Analysing the vibration signals is challenging when vibrations from other processes surrounding the ladle or converter may affect the measurements. Even do vibration measured on ladles and converters has been reported before, the installation of vibration sensors must be adapted to plant-specific conditions as the temperature of sensors’ mounting positions and surrounding processes create vibrations. Positions for mounting the accelerometer were found close enough to monitor the vibrations from the process but at the same time at a position where the heat from the steel melt did not overheat the sensors. The positions of the accelerometers did not negatively affect the logistic of the steel production or the regular working procedures at the steel plants. 

Different mathematical algorithms were used to evaluate the recorded signals as machine learning, wavelet transform and the commonly used Fourier change. A comparison of the results from these different algorithms with available process parameters shows that the vibrations of the process correlate to the process performance. The analysed results correlate to the regulated gas flow at the vacuum ladle treatment and deviations most likely caused by fewer working porous plugs. In the AOD converter, the refining steps may be followed as different vibration intensities. 

CLASS:PUBLIC
DTSTART:20230615T100000
DTEND:20230615T102000
END:VEVENT
END:VCALENDAR