Olfactory nuisances are an issue of growing concern considering that people's awareness about the effect of pollution on health and environment is increasing and that the perception of odor is related to a possible warning situation. A recent review concluded that odor assessment has to be faced by an integrated multi-tool strategy. In this paper we discuss the building of a model by means of a chemometric approach based on artificial neural networks known as Self Organizing Maps. These are applied to data collected by electronic nose continuous monitoring. The Self Organizing Map output was subjected to a second level clusterization by k-means algorithm. The cluster interpretation (i.e. air types classification in terms of “malodor”/“odor free” attributes) is achieved by crosslinking data produced by different instrumental and sensorial approaches, allowing us to establish the Frequency-Intensity-Duration odor characteristics for every identified air type. In order to elucidate our approach we focused our work on a four months survey at a residential site close to an integral cycle steel plant in Trieste (Italy). Odor Control Map proved to be a promising tool to provide valuable visualization support for following the dynamic evolution of the system with time. It allowed us to, for example, identify the relationships among sensor responses in different air types; follow the changes of air types with time. identify possible malodor sources; experimentally evaluate the frequency and duration of air types classified as malodorous. Furthermore, this first application highlighted the possible method improvements that have to be tested in different real environmental scenarios to obtain more robust and refined models. Considering that different annoyances (e.g., odor, noise, presence of specific chemical compounds) can cause possible synergistic health effects, Odor Control Map is a suitable tool to integrate data deriving from different and independent analysis/monitoring to obtain a more comprehensive knowledge on complex environmental phenomena involving dwellings close to industrial plants.
Odor control map: Self organizing map built from electronic nose signals and integrated by different instrumental and sensorial data to obtain an assessment tool for real environmental scenarios
Licen, Sabina
Writing – Review & Editing
;Briguglio, Sara Carmela;Barbieri, PierluigiSupervision
2018-01-01
Abstract
Olfactory nuisances are an issue of growing concern considering that people's awareness about the effect of pollution on health and environment is increasing and that the perception of odor is related to a possible warning situation. A recent review concluded that odor assessment has to be faced by an integrated multi-tool strategy. In this paper we discuss the building of a model by means of a chemometric approach based on artificial neural networks known as Self Organizing Maps. These are applied to data collected by electronic nose continuous monitoring. The Self Organizing Map output was subjected to a second level clusterization by k-means algorithm. The cluster interpretation (i.e. air types classification in terms of “malodor”/“odor free” attributes) is achieved by crosslinking data produced by different instrumental and sensorial approaches, allowing us to establish the Frequency-Intensity-Duration odor characteristics for every identified air type. In order to elucidate our approach we focused our work on a four months survey at a residential site close to an integral cycle steel plant in Trieste (Italy). Odor Control Map proved to be a promising tool to provide valuable visualization support for following the dynamic evolution of the system with time. It allowed us to, for example, identify the relationships among sensor responses in different air types; follow the changes of air types with time. identify possible malodor sources; experimentally evaluate the frequency and duration of air types classified as malodorous. Furthermore, this first application highlighted the possible method improvements that have to be tested in different real environmental scenarios to obtain more robust and refined models. Considering that different annoyances (e.g., odor, noise, presence of specific chemical compounds) can cause possible synergistic health effects, Odor Control Map is a suitable tool to integrate data deriving from different and independent analysis/monitoring to obtain a more comprehensive knowledge on complex environmental phenomena involving dwellings close to industrial plants.File | Dimensione | Formato | |
---|---|---|---|
1-s2.0-S0925400518304246-main.pdf
Accesso chiuso
Tipologia:
Documento in Versione Editoriale
Licenza:
Copyright Editore
Dimensione
1.94 MB
Formato
Adobe PDF
|
1.94 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
2921858_1-s2.0-S0925400518304246-main-PostPrint.pdf
accesso aperto
Descrizione: Post Print VQR3
Tipologia:
Bozza finale post-referaggio (post-print)
Licenza:
Digital Rights Management non definito
Dimensione
2.19 MB
Formato
Adobe PDF
|
2.19 MB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.