Abstract
Ultrasonic Machining (USM) has broad applicability in numerous industrial fields. Accurate capture of ultrasonic vibration signals is pivotal to the functioning of USM, making them areas of significant research interest. However, the nonlinear and non-stationary nature of the USM vibration signal makes it unsuitable for analysis with conventional methods such as Fast Fourier Transform. Despite current methodologies like Discrete Wavelet Transformation (DWT) yielding valuable insights, they involve manual parameter selection and could lead to sub-optimal results. This paper presents a novel method, using Variational Mode Decomposition (VMD) to automatically decompose USM vibration signals into intrinsic mode functions (IMFs). This method is complemented by Particle Swarm Optimization (PSO) algorithm to optimize the number of IMFs and penalty factor, with Sample Entropy (SE) serving as a fitness function. The innovative SE-PSO-VMD method is validated in ultrasonic metal welding and demonstrates a notable ability in predicting the pull force of welding materials with a high coefficient of determination R2 value of 0.78.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
