Mining Data of Noisy Signal Patterns in Recognition of Gasoline Bio-Based Additives using Electronic Nose

Abstract

The paper analyses the distorted data of an electronic nose in recognizing the gasoline bio-based additives. Different tools of data mining, such as the methods of data clustering, principal component analysis, wavelet transformation, support vector machine and random forest of decision trees are applied. A special stress is put on the robustness of signal processing systems to the noise distorting the registered sensor signals. A special denoising procedure based on application of discrete wavelet transformation has been proposed. This procedure enables to reduce the error rate of recognition in a significant way. The numerical results of experiments devoted to the recognition of different blends of gasoline have shown the superiority of support vector machine in a noisy environment of measurement.

A systematic bibliometric analysis of studies dealing with fuel-related e-nose applications

“…Table 5 lists the 22 articles describing the application of e-nose to fuel-related products published over the past ten years and selected on the basis of relevance to the present analysis (Aliaño-González et al , 2018a, 2018b; Amini and Hosseini-Golgoo, 2012; Bieganowski et al , 2018; Calle et al , 2020; Falatová et al , 2018, 2021; Ferreiro-González et al , 2016, 2017; Hong, 2018; Kumar et al , 2020; López et al , 2016; Mahmodi et al , 2019; Mumyakmaz and Karabacak, 2015; Nozza et al , 2016; Osowski and Siwek, 2017; Singh et al , 2016; Siqueira et al , 2018, 2019; Song et al , 2011; Vidigal et al , 2021; Wu et al , 2020). Among these papers, 10 (45.5%) were published in the periodical sensors, whereas the remainder were distributed over a diverse range of journals.…”

“…Among these papers, 10 (45.5%) were published in the periodical sensors, whereas the remainder were distributed over a diverse range of journals. A total of 15 articles (68.1%) were classified in the discrimination group (Aliaño-González et al , 2018a; Amini and Hosseini-Golgoo, 2012; Bieganowski et al , 2018; Calle et al , 2020; Falatová et al , 2018, 2021; Ferreiro-González et al , 2016, 2017; Kumar et al , 2020; Mahmodi et al , 2019; Osowski and Siwek, 2017; Singh et al , 2016; Siqueira et al , 2018, 2019; Song et al , 2011), five articles (22.7%) were classified in the prediction group (Hong, 2018; López et al , 2016; Mumyakmaz and Karabacak, 2015; Nozza et al , 2016; Vidigal et al , 2021) and the remaining two studies (Aliaño-González et al , 2018b; Wu et al , 2020) could be classified in both groups.…”

“…The full set of papers within the discrimination group used a range of statistical techniques to distinguish the e-nose profiles of the samples as follows: 9 used principal component analysis (PCA) (Aliaño-González et al , 2018b; Bieganowski et al , 2018; Falatová et al , 2018; Ferreiro-González et al , 2017; Kumar et al , 2020; Osowski and Siwek, 2017; Singh et al , 2016; Siqueira et al , 2018; Wu et al , 2020), 9 used linear discriminant analysis (LDA) (Aliaño-González et al , 2018a, 2018b; Amini and Hosseini-Golgoo, 2012; Calle et al , 2020; Falatová et al , 2018, 2021; Ferreiro-González et al , 2016, 2017; Mahmodi et al , 2019), 7 used artificial neural networks (ANN) (Amini and Hosseini-Golgoo, 2012; Bieganowski et al , 2018; Kumar et al , 2020; Singh et al , 2016; Siqueira et al , 2018; Song et al , 2011; Wu et al , 2020), 7 used hierarchical clustering analysis (HCA) (Aliaño-González et al , 2018a, 2018b; Calle et al , 2020; Falatová et al , 2018, 2021; Ferreiro-González et al , 2016, 2017), 3 used quadratic discriminant analysis (QDA) (Mahmodi et al , 2019; Siqueira et al , 2018, 2019), 3 used support vector machine (SVM) (Mahmodi et al , 2019; Osowski and Siwek, 2017; Song et al , 2011) and 3 used stochastic analysis (Siqueira et al , 2018, 2019; Vidigal et al , 2021). Other methods mentioned were autoregressive moving average with exogenous inputs (ARMAX) (Amini and Hosseini-Golgoo, 2012), wavelet analysis (Osowski and Siwek, 2017) and random forest analyses (Osowski and Siwek, 2017).…”

“…A total of 15 articles (68.1%) were classified in the discrimination group (Aliaño-Gonz alez et al, 2018a; Amini and Hosseini-Golgoo, 2012; Bieganowski et al, 2018;Calle et al, 2020;Falatov a et al, 2018Falatov a et al, , 2021Ferreiro-Gonz alez et al, 2016Mahmodi et al, 2019 stochastic analysis (Siqueira et al, 2018(Siqueira et al, , 2019Vidigal et al, 2021). Other methods mentioned were autoregressive moving average with exogenous inputs (ARMAX) (Amini and Hosseini-Golgoo, 2012), wavelet analysis (Osowski and Siwek, 2017) and random forest analyses (Osowski and Siwek, 2017). Within all articles classified in the prediction group, four papers predicted the concentration of gases and volatile vapors using simple linear regression (Wu et al, 2020), multiple linear regression (Vidigal et al, 2021), mass balance model/ confidence intervals (Nozza et al, 2016) or ANN (Mumyakmaz and Karabacak, 2015).…”

Quantification of Acrylonitrile Butadiene Styrene Odor Intensity Based on a Novel Odor Assessment System With a Sensor Array

“…On the other hand, the pattern recognition methods are employed to classify the signals representing the smell features. Osowski and Siwek [28] utilized the electronic nose to identify the distorted data of biological additives in the gasoline by means of principal component analysis (PCA), wavelet transform, support vector machine (SVM), etc., summing up the advantages of SVM in reducing errors. In the past few years, the olfactory bionic technology has progressed in the evaluation of the grades of paraffin [29], air quality [30], and the flavors of beverage [31]- [33].…”