Solid–liquid mixing is a core operation in many manufacturing processes in the food, cosmetics, pharmaceutical and chemical industries. This work aims to develop an accurate and reliable sensing methodology using passive acoustic emission (PAE) coupled with supervised machine learning (ML) algorithms, to allow identifying and predicting solid–liquid suspension state. Using PAE in process monitoring is beneficial because it is affordable, sensitive, non-intrusive, and suitable for on-line applications. PAE equipment includes a piezoelectric sensor, placed in contact with the system, an amplifier, a filter, an oscilloscope to record the signal and a computer. Experiments were carried out in a fully baffled, flat bottom glass vessel equipped with a PBT impeller. Acoustic signals were recorded with sampling frequency of 750 kHz, impeller speed range 50–1000 rpm and varying solid features, i.e., particle size (dp range 0.250–6 mm), solid loading and solid density (acryl-glass particles). For each classification run, sampled data were pre-processed using Fast Fourier Transform (FFT) to reveal any detailed spectral characteristics of the signal in the frequency domain. Spectra have been filtered and then reduced by selecting the highest variance frequencies. As labelling, established optical measurements were used to classify the acoustic frequency spectra. The frequency data set has been split in training (60%), cross validation (20%) and test (20%) sets and were used, respectively, to build the model, identify the best model parameters (optimisation step), and finally to check the accuracy (test step). The developed technique has shown excellent results in recognizing spectra corresponding to different classes with observed accuracy greater than 99.72%.

Identification of suspension state using passive acoustic emission and machine learning in a solid–liquid mixing system

Brunazzi E.
Ultimo
Writing – Review & Editing
2022-01-01

Abstract

Solid–liquid mixing is a core operation in many manufacturing processes in the food, cosmetics, pharmaceutical and chemical industries. This work aims to develop an accurate and reliable sensing methodology using passive acoustic emission (PAE) coupled with supervised machine learning (ML) algorithms, to allow identifying and predicting solid–liquid suspension state. Using PAE in process monitoring is beneficial because it is affordable, sensitive, non-intrusive, and suitable for on-line applications. PAE equipment includes a piezoelectric sensor, placed in contact with the system, an amplifier, a filter, an oscilloscope to record the signal and a computer. Experiments were carried out in a fully baffled, flat bottom glass vessel equipped with a PBT impeller. Acoustic signals were recorded with sampling frequency of 750 kHz, impeller speed range 50–1000 rpm and varying solid features, i.e., particle size (dp range 0.250–6 mm), solid loading and solid density (acryl-glass particles). For each classification run, sampled data were pre-processed using Fast Fourier Transform (FFT) to reveal any detailed spectral characteristics of the signal in the frequency domain. Spectra have been filtered and then reduced by selecting the highest variance frequencies. As labelling, established optical measurements were used to classify the acoustic frequency spectra. The frequency data set has been split in training (60%), cross validation (20%) and test (20%) sets and were used, respectively, to build the model, identify the best model parameters (optimisation step), and finally to check the accuracy (test step). The developed technique has shown excellent results in recognizing spectra corresponding to different classes with observed accuracy greater than 99.72%.
2022
Rossi, A.; Alberini, F.; Brunazzi, E.
File in questo prodotto:
File Dimensione Formato  
ChERD2022-Rossi-etal.pdf

solo utenti autorizzati

Tipologia: Versione finale editoriale
Licenza: NON PUBBLICO - Accesso privato/ristretto
Dimensione 2.25 MB
Formato Adobe PDF
2.25 MB Adobe PDF   Visualizza/Apri   Richiedi una copia
Cherd2022_Rossi_etal_PrePrint.pdf

Open Access dal 09/01/2024

Descrizione: versione post-print
Tipologia: Documento in Post-print
Licenza: Creative commons
Dimensione 2.66 MB
Formato Adobe PDF
2.66 MB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11568/1120289
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 4
  • ???jsp.display-item.citation.isi??? 4
social impact