Title: Trimming process sheet thickness characterisation with mel-frequency cepstral coefficients and an artificial neural network

Authors: Tushar Y. Badgujar; Harshal A. Chavan; Shubham R. Suryawanshi; Vijay P. Wani

Addresses: Mechanical Engineering Department, Late G. N. Sapkal College of Engineering, Affiliated to Savitribai Phule Pune University (SPPU), Nashik, Maharashtra, 422213, India ' Mechanical Engineering Department, Mumbai Educational Trust's Institute of Engineering, Affiliated to Savitribai Phule Pune University, Nashik, Maharashtra, 422003, India ' Mechanical Engineering Department, Mumbai Educational Trust's Institute of Engineering, Affiliated to Savitribai Phule Pune University, Nashik, Maharashtra, 422003, India ' Mechanical Engineering Department, Mumbai Educational Trust's Institute of Engineering, Affiliated to Savitribai Phule Pune University, Nashik, Maharashtra, 422003, India

Abstract: Sheet metal trimming is one of the important manufacturing processes in modern industry. Shear stress in the trimming process has a significant impact on material behaviour, tool wear, and product quality. When other parameters remain unchanged, the thickness of the sheet metal determines the shear stress, so it becomes important to maintain thickness within a predetermined range. The present study aims to monitor sheet thickness variation in the trimming process online using the wavelet transform as a signal filter, the mel-frequency cepstral coefficients (MFCC's) as a feature, and artificial neural networks (ANNs) as a classifier. Experimental results showed accuracy of 91.63% and 89.5% with pre-recorded acoustic signals and experimental trials, respectively. The proposed online sheet thickness monitoring system has the potential to improve productivity by reducing inspection time and providing insight into shear stress.

Keywords: sheet metal trimming; shear stress; MFCC; mel-frequency cepstral coefficient; ANN; artificial neural networks; online monitoring.

DOI: 10.1504/IJMMS.2024.144281

International Journal of Mechatronics and Manufacturing Systems, 2024 Vol.17 No.4, pp.370 - 386

Received: 15 Apr 2024
Accepted: 27 Oct 2024
Published online: 04 Feb 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article