Analysis And Research On Mould Friction Signal Based On Wavelet Theory

The mould friction is an important parameter reflecting the initial shell and mould lubrication conditions. The research of mould friction is very important to the optimization and developing of continuous casting. With the developing of highly active continuous casting technology, questions like strand quality and breaking are highlighted increasingly. So the understanding and controlling of lubrication in mould become a more and more important part of continuous casting research.At present, certain achievements of the study of mould friction characteristics has been made in the time domain which provide important guidance for the production. But there are some difficulties to analyze the mould friction fluctuation only in the time domain due to many impact factors of the mould friction. As a new method to research the mould friction, wavelet analysis can analyze time-varying mould friction signals with the multi-scale analysis and investigate the time-frequency characteristics of the mould friction signals intensively which has important theoretical and practical significance to help to reveal the nature and mechanism of mould friction behavior and develop the continuous casting technology.This article aims to research the mould friction signals intensively based on the wavelet theory. Hence, the time-frequency characteristics and the wavelet entropy characteristics of the mould friction signals were researched through the study of the mould friction measurement theory and wavelet theory.(1) The study of the mould friction measurement theory helped to realize the measurement of the mould friction which built a solid foundation for the application of wavelet theory to research the mould friction signals.(2) With the study of the wavelet theory, the traditional screening method was used to select the suitable wavelet function to analyze the mould friction signals based on the signals features.(3) The wavelet transform was applied to the typical mould friction signals and the wavelet time-frequency maps of the mould friction clearly characterized the time-frequency information at sinusoidal and nonsinusoidal oscillation stage, casting speed increase and decrease stage, casting speed fluctuation stage and breakout stage. Wavelet transform can effectively analyze the mould friction signal which has potential applications for the mould friction research.(4) The establishment of the mould friction wavelet entropy analysis model combined the wavelet local analysis ability and entropy statistical properties. This realized the detailed analysis of the mould friction signals. Also, the analysis results were compared with the mould friction RMS.(5) According to the analysis results of the mould friction wavelet entropy, the fluctuation characteristics of the mould friction were obtained when abnormal production occurred such as acute mould level fluctuations, submerged entry nozzle broken, breakout and so on. The statistical analysis of 15 groups mould friction wavelet entropy when abnormal casting were carried out and the results showed that the mould friction wavelet entropy was quite sensitive to the change of the lubrication condition in the mould which has broad application prospect in the field of continuous casting.