Adaptive decomposition and feature identification of milling chatter signals based on ITSLWOA-VMD-MPE

In machining systems, chatter is a harmful dynamic instability phenomenon that can cause high-frequency oscillations between the tool and the workpiece, exacerbate tool wear, and affect machining quality. To achieve efficient and accurate vibration monitoring, this study proposes an integrated optimized variational mode decomposition (VMD) and multi-scale permutation entropy (MPE) milling vibration signal analysis method—ITSLWOA-VMD-MPE. This method combines the Improved Whale Optimization Algorithm (ITSLWOA) to achieve adaptive signal decomposition and utilizes MPE for chatter state recognition. Firstly, Tent chaotic mapping, sine adaptive weighting, and Levy flight optimization WOA are introduced to enhance its search efficiency and global exploration capability. Secondly, ITSLWOA is used to adaptively optimize VMD parameters and accurately extract vibration signals during the milling process. Finally, combining MPE to quantify the evolution of chatter states can improve the accuracy of chatter identification. Simulation and experimental results show that this method can efficiently and accurately decompose milling signals, enhance the ability to extract chatter features, and effectively improve the robustness of chatter recognition.