超声辅助加工系统的刀具状态自感知算法

超声波振动台内含压电材料，可以拾取切削过程产生的振动信号，实现不借助外部传感器刀具工作状态的自感知。为了从刀具振动信号中获取有效信息，该文提出一种基于经验模态分解的时频域重构算法。首先，采用经验模态分解算法将原始信号分解，得到多个固有模态函数分量和残差分量；其次，计算原始信号与各分量之间的时频域互相关系数；再次，归一化时频域互相关系数作为权重值，将固有模态函数分量和残差进行重构；最后，通过数值仿真和超声辅助加工实验，验证了基于经验模态分解的时频域重构算法的去噪性能，提取了信噪比为5.03 dB的目标信号，从而实现了超声辅助加工系统的自感知功能。

Tool state self-sensing algorithm used for ultrasonic assisted machining system

Tool condition monitoring technology is the key to realize intelligent manufacturing, and it is of great significance to ensure processing quality and improve processing efficiency. Through superimposing high-frequency ultrasonic vibration on the tool or workpiece, ultrasonic assisted machining increases the cutting efficiency and improves the surface quality of the workpiece. Due to containing piezoelectric material, the ultrasonic vibration table can pick up the vibration signal generated by the cutting process, and realize the self-sensing of the working state of the tool without additional sensors. In order to extract the active ingredients from the tool vibration signal, a time and frequency domain reconstruction algorithm based on empirical mode decomposition (TF-EMD) is proposed. Firstly, the original signal is decomposed into multiple intrinsic mode function (IMF) components and the residual by empirical mode decomposition algorithm. Secondly, the cross-correlation coefficients are calculated between the original signal and decomposed results in both time domain and frequency domain. Thirdly, the weighted factors are obtained by normalizing cross-correlation coefficients, and the IMF components and residual are reconstructed through the obtained weighted factors. Finally, numerical simulation and ultrasonic assisted machining experiment are carried out to verify the denoising performance of the TF-EMD algorithm. The signal with a signal-to-noise ratio of 5.03dB is extracted, thus the self-sensing of the ultrasonic assisted machining system is realized.