Simulation Identification of Milling Chatter Based on Improved VMD Algorithm

颤振是机床铣削加工中的关键问题，严重影响加工效率与工件质量，因此早期颤振检测具有重要意义。本文提出一种基于优化变分模态分解与功率谱熵结合的铣削颤振识别方法。针对变分模态分解参数选择难题，采用麻雀搜索算法与最小包络熵相结合的自适应优化策略，实现参数高效寻优；基于分解信号能量比筛选本征模态分量，重构信号以去除噪声干扰；引入功率谱熵作为颤振识别指标，提取仿真信号特征并实现颤振状态的精准识别。实验结果表明，该方法能够有效检测颤振状态。Chatter is a critical issue in machine tool milling, seriously affecting machining efficiency and workpiece quality, so early chatter detection is of great significance. This paper proposes a milling chatter identification method based on optimized variational mode decomposition (VMD) and power spectrum entropy (PSE). To address the challenge of VMD parameters selection, an adaptive optimization strategy combining sparrow search algorithm (SSA) and minimum envelope entropy (EE) is adopted to achieve efficient parameters optimization. Based on the energy ratio of the decomposed signals, intrinsic mode functions (IMFs) are selected to reconstruct the signals and eliminate noise interference. PSE is introduced as a chatter identification index to extract characteristics of simulated signal and achieve accurate identification of chatter states. Experimental results demonstrate that the proposed method can effectively detect chatter.