Denoising Method of Vibration Signal for CNC Machine Tool Multi-Component Feed System Based on Joint Analysis Method

Abstract

The feed system is an important part of CNC machine tools, and its condition monitoring is mostly based on vibration signals measured by sensors. To remove the noise mixed in the signal, a joint denoising method based on variational mode decomposition (VMD), simple correlation analysis (SCA) and translation invariant wavelet (TIW) denoising is proposed in this paper. Firstly, the VMD parameters are adjusted adaptively by the Aquila Optimizer (AO) to meet the demand of signal decomposition from different components of feed system. Then, the intrinsic mode functions (IMFs) obtained after VMD processing are reconstructed based on correlation analysis to eliminate irrelevant information. Finally, the reconstructed signal is denoised by translation invariant wavelet to obtain the denoised signal. The feasibility and universality of the joint analysis method are verified by the denoising test of simulation signals and measured signals from a certain machine feed system. The results show that the joint analysis method has better denoising effect than some general denoising method, and it can satisfy the denoising requirement when processing signals from different components. Besides, the proposed method has the ability of adaptive denoising. Compared with other optimization algorithms, AO algorithm has better optimization accuracy and efficiency, which can provide good support for the universality of the joint analysis denoising method.

Keywords

• Aquila optimizer
• machine tool
• multi-component feed system
• signal denoising
• variational mode decomposition