The aim of this study was to develop and test a new fuzzy logic model for monitoring the udder health status (HS) of goats. The model evaluated, as input variables, the milk electrical conductivity (EC) signal, acquired on-line for each gland by a dedicated sensor, the bandwidth length and the frequency and amplitude of the first main peak of the Fourier frequency spectrum of the recorded milk EC signal. Two foremilk gland samples were collected from eight Saanen goats for six months at morning milking (lactation stages (LS): 0–60 Days In Milking (DIM); 61–120 DIM; 121–180 DIM), for a total of 5592 samples. Bacteriological analyses and somatic cell counts (SCC) were used to define the HS of the glands. With negative bacteriological analyses and SCC < 1,000,000 cells/mL, glands were classified as healthy. When bacteriological analyses were positive or showed a SCC > 1,000,000 cells/mL, glands were classified as not healthy (NH). For each EC signal, an estimated EC value was calculated and a relative deviation was obtained. Furthermore, the Fourier frequency spectrum was evaluated and bandwidth length, frequency and amplitude of the first main peak were identified. Before using these indexes as input variables of the fuzzy logic model a linear mixed-effects model was developed to evaluate the acquired data considering the HS, LS and LS × HS as explanatory variables. Results showed that performance of a fuzzy logic model, in the monitoring of mammary gland HS, could be improved by the use of EC indexes derived from the Fourier frequency spectra of gland milk EC signals recorded by on-line EC sensors.

Improved fuzzy logic system to evaluate milk electrical conductivity signals from on-line sensors to monitor dairy goat mastitis / M. Zaninelli, F.M. Tangorra, A. Costa, L. Rossi, V. Dell’Orto, G. Savoini. - In: SENSORS. - ISSN 1424-8220. - 16(2016 Jul), pp. 1079.1-1079.18.

Improved fuzzy logic system to evaluate milk electrical conductivity signals from on-line sensors to monitor dairy goat mastitis

F.M. Tangorra
;
A. Costa
;
L. Rossi
;
V. Dell’Orto
Penultimo
;
G. Savoini
Ultimo
2016

Abstract

The aim of this study was to develop and test a new fuzzy logic model for monitoring the udder health status (HS) of goats. The model evaluated, as input variables, the milk electrical conductivity (EC) signal, acquired on-line for each gland by a dedicated sensor, the bandwidth length and the frequency and amplitude of the first main peak of the Fourier frequency spectrum of the recorded milk EC signal. Two foremilk gland samples were collected from eight Saanen goats for six months at morning milking (lactation stages (LS): 0–60 Days In Milking (DIM); 61–120 DIM; 121–180 DIM), for a total of 5592 samples. Bacteriological analyses and somatic cell counts (SCC) were used to define the HS of the glands. With negative bacteriological analyses and SCC < 1,000,000 cells/mL, glands were classified as healthy. When bacteriological analyses were positive or showed a SCC > 1,000,000 cells/mL, glands were classified as not healthy (NH). For each EC signal, an estimated EC value was calculated and a relative deviation was obtained. Furthermore, the Fourier frequency spectrum was evaluated and bandwidth length, frequency and amplitude of the first main peak were identified. Before using these indexes as input variables of the fuzzy logic model a linear mixed-effects model was developed to evaluate the acquired data considering the HS, LS and LS × HS as explanatory variables. Results showed that performance of a fuzzy logic model, in the monitoring of mammary gland HS, could be improved by the use of EC indexes derived from the Fourier frequency spectra of gland milk EC signals recorded by on-line EC sensors.
fuzzy logic; electrical conductivity; mastitis; dairy goats
Settore AGR/09 - Meccanica Agraria
Settore AGR/10 - Costruzioni Rurali e Territorio Agroforestale
Settore AGR/18 - Nutrizione e Alimentazione Animale
Article (author)
File in questo prodotto:
File Dimensione Formato  
sensors-16-01079 (1).pdf

accesso aperto

Tipologia: Publisher's version/PDF
Dimensione 813.45 kB
Formato Adobe PDF
813.45 kB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

Caricamento pubblicazioni consigliate

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/2434/421097
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 10
  • ???jsp.display-item.citation.isi??? 10
social impact