Multi-scale shell vibration signal analysis of ball mill in grinding process

Abstract

Component decomposition and physical interpretation of shell vibration signal of ball mill in grinding process is very difficult. Aim to this problem, several multi-scales decomposition algorithms are used to analyze shell vibration signal in this paper. Empirical mode decomposition, ensemble empirical mode decomposition and Hilbert vibration decomposition algorithms decompose shell vibration signal to a series of multi-scale sub-signals. Correlation coefficients between these sub-signals and original vibration signal were calculated. Power density spectrums of different multi-scale sub-signals used also to analyze and explain the shell vibration signal. Results show that different decomposition algorithms have different advantages and shortcomings. The further research direction is that how to fuse these sub-signals and model load parameters inside the ball mill.