Multi-signal comparison and analysis of tool wear based on EMD energy decomposition method

This paper proposes a comparative analysis method based on EMD energy decomposition for detecting the amount of tool wear during milling and to achieve online monitoring of tool wear status in order to ensure the quality and accuracy of products machined by machine tools and to lower the risk of damage to products and machine tools due to tool wear or breakage. Three separate sensors were employed for the planned milling experiment to gather current signal, vibration signal and sound signal respectively. The varied signal energies were then processed using the EMD energy decomposition technique, and the relationship between the outcomes and tool wear was examined.The characteristic frequency bands eventually resulted in the acquisition of six intrinsic mode function (IMF) components for three different wear periods and the energy calculation of each IMF component because the acquired signals were non-stationary and non-linear.. Finally, analysis and comparison of the overall energy distribution and trends of each roup of signals was performed. According to the experimental findings, under the identical machining settings, the vibration signals based on the EMD energy decomposition have a high correlation with tool wear, followed by the current signals, while the sound signals do not show a significant correlation with tool wear. The results of this paper have some significance for tool wear monitoring.