Determination of Odour Interactions of Three-Component Gas Mixtures Using an Electronic Nose

Porous silicon opto-acoustic detector for ternary gas mixture

The use of porous phoxonic crystals with coupled optical and acoustic response has been proposed as a sensing device. Due to the porous nature of the crystal, each layer of the structure can be connected to the environment. As the optical and acoustic performances of the phoxonic crystal change when a gas permeates the pores due to modifications of the effective refractive index and density of the system, it results that these structures are suitable platforms for the detection of gases. The sensor designed following these premises can detect the composition of ternary gas mixtures through optical measurements, while an acoustic wave induces a structural oscillation. The amplified acoustic wave produces a mechanical deformation of the crystal layers that is maximized in the center a resonant microcavity. Therefore, under such experimental conditions, the sensitivity of the optical response is not only due to the optical property changes caused by the gas mixture in contact with the porous structure but also to changes in the mechanical deformations due to modifications of the acoustic properties. In this work, we discuss the device theoretical behavior as a multiparameter sensor that distinguishes the components and concentrations of a ternary gas mixture through the transfer matrix method. For a prototype combination of CO2-Air-CH4 mixture, the estimated resolution of the proposed device fabricated in porous silicon can be has high as 0.05% (500 ppm) in the concentration of each individual species.

Characterization of odour emissions in a wastewater treatment plant using a drone-based chemical sensor system

Conventionally, odours emitted by different sources present in wastewater treatment plants (WWTPs) are measured by dynamic olfactometry, where a human panel sniffs and analyzes air bags collected from the plant. Although the method is considered the gold standard, the process is costly, slow, and infrequent, which does not allow operators to quickly identify and respond to problems. To better monitor and map WWTP odour emissions, here we propose a small rotary-wing drone equipped with a lightweight (1.3-kg) electronic nose. The "sniffing drone" sucks in air via a ten-meter (33-foot) tube and delivers it to a sensor chamber where it is analyzed in real-time by an array of 21 gas sensors. From the sensor signals, machine learning (ML) algorithms predict the odour concentration that a human panel using the EN13725 methodology would report. To calibrate and validate the predictive models, the drone also carries a remotely controlled sampling device (compliant with EN13725:2022) to collect sample air in bags for post-flight dynamic olfactometry. The feasibility of the proposed system is assessed in a WWTP in Spain through several measurement campaigns covering diverse operating regimes of the plant and meteorological conditions. We demonstrate that training the ML algorithms with dynamic (transient) sensor signals measured in flight conditions leads to better performance than the traditional approach of using steady-state signals measured in the lab via controlled exposures to odour bags. The comparison of the electronic nose predictions with dynamic olfactometry measurements indicates a negligible bias between the two measurement techniques and 95 % limits of agreement within a factor of four. This apparently large disagreement, partly caused by the high uncertainty of olfactometric measurements (typically a factor of two), is more than offset by the immediacy of the predictions and the practical advantages of using a drone-based system.

Hydrophobic and Rapid-Response Sensor Inks: Array-Based Fingerprinting of Perfumes

Counterfeited perfumes mixed with inexpensive additives for commercial purposes pose a great threat to cosmetic market competition and human health. Herein, a 24-element, solid-state colorimetric sensor array employing chemo-responsive dye inks for accurate discrimination of a variety of fragrance bases and "sniffing out" real perfumes from adulterated samples was first reported. The physiochemical robustness and gas response kinetics of the sensor array were optimized with the streamlined design of the channel geometry and hydrophobic modification of the sensor substrate. A unique and distinguishable color change profile was obtained within 2 min exposure of diluted vapor that enabled clear fingerprinting of chemically similar perfume samples. Four commercial perfume products were successfully distinguished and categorized according to their similarity to relevant perfume bases using chemometric methods including hierarchical clustering and principal component analysis. The sensor array also allows the discrimination of ethanol-diluted fragrance bases from the pristine sample, revealing its potential for quality assurance of perfumes and other cosmetics. Such easy-to-use, disposable, and miniaturized chemical sensing detectors therefore prove exceptionally valuable for fast analysis of luxuries such as perfumes and other industrial products with complex chemical compositions.

Metal loaded nano-carbon gas sensor array for pollutant detection. NANOTECHNOLOGY 2022;33:195501. [PMID: 35073524 DOI: 10.1088/1361-6528/ac4e43]

Many research works report a sensitive detection of a wide variety of gas species. However, their in-lab detection is usually performed by using single gases and, therefore, selectivity often remains an unsolved issue. This paper reports a four-sensor array employing different nano-carbon sensitive layers (bare graphene, SnO₂@Graphene, WO₃@Graphene, and Au@CNTs). The different gas-sensitive films were characterised via several techniques such as FESEM, TEM, and Raman. First, an extensive study was performed to detect isolated NO₂, CO₂, and NH₃molecules, unravelling the sensing mechanism at the operating temperatures applied. Besides, the effect of the ambient moisture was also evaluated. Afterwards, a model for target gas identification and concentration prediction was developed. Indeed, the sensor array was used in mixtures of NO₂and CO₂for studying the cross-sensitivity and developing a calibration model. As a result, the NO₂detection with different background levels of CO₂was achieved with anR²of 0.987 and an RMSE of about 22 ppb.

Noninvasive real-time breath test for controlling hormonal background of the human body: detection of serotonin and melatonin with quantum point-contact sensors

Significant progress in development of noninvasive diagnostic tools based on breath analysis can be expected if one employs a real-time detection method based on finding a spectral breath profile which would contain some energy characteristics of the analyzed gas mixture. Using the fundamental energy parameters of a quantum system, it is possible to determine with a high accuracy its quantitative and qualitative composition. Among the most efficient tools to measure energy characteristics of quantum systems are sensors based on Yanson point contacts. This paper reports the results of serotonin and melatonin detection as an example of testing the human hormonal background with point-contact sensors, which have already demonstrated their high efficiency in detecting carcinogenic strains ofHelicobacter pyloriand selective detection of complex gas mixtures. When comparing the values of serotonin and melatonin with the characteristic parameters of the spectral profile of the exhaled breath of each patient, high correlation dependences of the concentration of serotonin and melatonin with a number of characteristic parameters of the response curve of the point-contact sensor were found. The performed correlation analysis was complemented with the regression analysis. As a result, empiric regression relations were proposed to realize in practice the new non-invasive breath test for evaluation of the human hormonal background. Registration of the patient's breath profile using point-contact sensors makes it possible to easily monitor the dynamics of changes in the human hormonal background and perform a quantitative evaluation of serotonin and melatonin levels in the human body in real time without invasive interventions (blood collection) and expensive equipment or reagents.

Odor-Based Nanomechanical Discrimination of Fuel Oils Using a Single Type of Designed Nanoparticles with Nonlinear Viscoelasticity

Odors are one of the most diverse and complicated gaseous mixtures so that their discrimination is challenging yet attractive because of the rich information about their origin. The more similar the properties of odors are, the more difficult the discrimination becomes. The practical applications, however, often demand such discrimination, especially with a compact sensing platform. In this paper, we show that a nanomaterial designed for a specific type of odors can clearly discriminate them even with a single nanomechanical sensing channel. Fuel oils and their mixture are used as a model target that has similar chemical properties but different compositions mainly consisting of paraffinic, olefinic, naphthenic, and aromatic hydrocarbons. We demonstrate using octadecyl functionalized silica-titania nanoparticles that the difference in the compositions is successfully picked up based on their high affinity for the aliphatic hydrocarbons and alkyl chain length dependent nonlinear viscoelastic behavior. Such a properly designed material is proved to derive sufficient information from a series of analytes to discriminate them even with a single sensing element. This approach provides a guideline to prepare various sensors whose response properties are distinct and optimized depending on applications.

On the Use of Pulsed UV or Visible Light Activated Gas Sensing of Reducing and Oxidising Species with WO3 and WS2 Nanomaterials

This paper presents a methodology to quantify oxidizing and reducing gases using n-type and p-type chemiresistive sensors, respectively. Low temperature sensor heating with pulsed UV or visible light modulation is used together with the application of the fast Fourier transform (FFT) to extract sensor response features. These features are further processed via principal component analysis (PCA) and principal component regression (PCR) for achieving gas discrimination and building concentration prediction models with R² values up to 98% and RMSE values as low as 5% for the total gas concentration range studied. UV and visible light were used to study the influence of the light wavelength in the prediction model performance. We demonstrate that n-type and p-type sensors need to be used together for achieving good quantification of oxidizing and reducing species, respectively, since the semiconductor type defines the prediction model’s effectiveness towards an oxidizing or reducing gas. The presented method reduces considerably the total time needed to quantify the gas concentration compared with the results obtained in a previous work. The use of visible light LEDs for performing pulsed light modulation enhances system performance and considerably reduces cost in comparison to previously reported UV light-based approaches.

Recent Developments in Smart Food Packaging Focused on Biobased and Biodegradable Polymers

Food packaging has a crucial function in the modern food industry. New food packaging technologies seek to meet consumers and industrial's demands. Changes related to food production, sale practices and consumers' lifestyles, along with environmental awareness and the advance in new areas of knowledge (such as nanotechnology or biotechnology), act as driving forces to develop smart packages that can extend food shelf-life, keeping and supervising their innocuousness and quality and also taking care of the environment. This review describes the main concepts and types of active and intelligent food packaging, focusing on recent progress and new trends using biodegradable and biobased polymers. Numerous studies show the great possibilities of these materials. Future research needs to focus on some important aspects such as possibilities to scale-up, costs, regulatory aspects, and consumers' acceptance, to make these systems commercially viable.

Drift Compensation on Massive Online Electronic-Nose Responses

Gas sensor drift is an important issue of electronic nose (E-nose) systems. This study follows this concern under the condition that requires an instant drift compensation with massive online E-nose responses. Recently, an active learning paradigm has been introduced to such condition. However, it does not consider the “noisy label” problem caused by the unreliability of its labeling process in real applications. Thus, we have proposed a class-label appraisal methodology and associated active learning framework to assess and correct the noisy labels. To evaluate the performance of the proposed methodologies, we used the datasets from two E-nose systems. The experimental results show that the proposed methodology helps the E-noses achieve higher accuracy with lower computation than the reference methods do. Finally, we can conclude that the proposed class-label appraisal mechanism is an effective means of enhancing the robustness of active learning-based E-nose drift compensation.

Analysis of volatile compounds causing undesirable odors in a polypropylene - high-density polyethylene recycled plastic resin with solid-phase microextraction

Although recycled plastics provide a low-cost and environmentally friendly alternative for many applications, their desirability is significantly limited by the presence of unpleasant odors from volatile organic compounds (VOCs). In this work, a headspace solid-phase microextraction (HS-SPME) coupled with gas chromatography-mass spectrometry (GC-MS) method was optimized to analyze volatile compounds from an odorous recycled plastic resin which was roughly composed of 85-90% polypropylene (PP) and 15-10% high-density polyethylene (HDPE). A large variety of aliphatic hydrocarbons and 13 additive residues were detected. Statistical tools were employed to screen the VOCs and successfully identified three components, i.e., 2,4-dimethyl-heptane, 4-methyl-octane and octamethylcyclotetrasiloxane (D₄), which were significantly related to the odor intensity of the recycled plastic resin (p-values < 0.05). 2,4-Dimethyl-heptane has a strong, pungent plastic smell, which is very similar to the odor of the recycled resin. It is identified as a major source of the odor. Past relevant research has not been able to establish a direct link between an odorous compound and the undesirable odor of recycled plastic until now. 4-Methyl-octane was highly corelated to 2,4-dimethyl-heptane and somewhat contributed to the odor. D₄ does not have an odor, but it may serve as an indicator of some odorous residues from personal care products.

A systematic study on the odorants characterization and evaluation in a plain reservoir with wetlands ecosystem

The odor problems in plain reservoirs are more complex compared to valley reservoirs and ground water reservoirs. Just as YL Reservoir with wetlands ecosystem in Lixiahe Plain, Jiangsu Province has been suffering from complex odors, however, the odorants were unclear. In this study, a systematic study on odorants characterization and evaluation of plain YL reservoir was accomplished. Totally musty, septic, fishy, chemical and grassy odors were first identified simultaneously, twenty-four odorants were identified correspondingly. According to odor activity value ranking, 2-methylisoborneol and geosmin, with odor activity values of 14-18 and 2.5-3.8, were major musty odorants, while bis(2-chloroisopropyl) ether, dimethyl disulfide, dimethyl trisulfide, pentanethiol and indole, with odor activity values of 15.3-18.8, 1-1.3, 1.5-2.3, 2-3 and 0.7-0.9, were major septic compounds. Fishy and grassy odors were associated with 2,4-decadienal, hexanal, nonanal, decanal, benzaldehyde and β-cyclocitral, while chemical odor was related to indane, eucalyptol, 2-nitrophenol, 2-methylphenol, tetramethylpyrazine, 1,4-dichlorobenzene, p-xylene and ethylbenzene. By reconstituting tests, 98, 95, 88, 85 and 81 % of musty, septic, fishy, chemical and grassy characteristics in source water could be explained. It was notable that wetlands ecosystem has almost no effect on odor removal. This study would offer more understandings and supports for odors's control in such source waters.

A Smart Rig for Calibration of Gas Sensor Nodes

Electrochemical gas sensors require regular maintenance to check and secure proper functioning. Standard procedures usually involve testing and recalibration of the sensors, for which working environments are needed. Periodic calibration is therefore necessary to ensure reliable and accurate measurements. This paper proposes a dedicated smart calibration rig with a set of novel features enabling simultaneous calibration of multiple sensors. The proposed calibration rig system comprises a gas mixing system, temperature control system, a test chamber, and a process-control PC that controls all calibration phases. The calibration process is automated by a LabVIEW-based platform that controls the calibration environment for the sensor nodes, logs sensor data, and best fit equation based on interpolation for every sensor on the node and uploads it to the sensor node for next deployments. The communication between the PC and the sensor nodes is performed using the same IEEE 802.15.4 (ZigBee) protocol that the nodes also use in field deployment for air quality measurement. The results presented demonstrate the effectiveness of the sensors calibration rig.

Integrated model for estimating odor emissions from civil wastewater treatment plants

The objective of this research project was the design and development of an integrated model for odor emission estimation in wastewater treatment plants. The SMAT's plant, the largest wastewater treatment facility in Italy, was used as a case study. This article reports the results of the characterization phase that led to the definition and design of the proposed conceptual model for odor emission estimation. In this phase, concentrations of odor chemical tracers (VOC, H₂S, NH₃) and odor concentrations were monitored repeatedly. VOC screening with GC-MS analysis was also performed. VOC concentrations showed significant variability in space and magnitude. NH₃ and H₂S were also detected at considerable concentrations. Results were elaborated to define a spatially variable linear relationship between the sum of odor activity values (SOAV) and odor concentrations. Based on the results, a conceptual operational model was presented and discussed. The proposed system is composed by a network of continuous measurement stations, a set of algorithms for data elaboration and synchronization, and emission dispersion modeling with the application of Lagrangian atmospheric models.

Simultaneous Removal of Hexane and Ethanol from Air in a Biotrickling Filter—Process Performance and Monitoring Using Electronic Nose

Biofiltration is a well-accepted method for the removal of malodorous compounds from air streams. Interestingly, the mechanisms underlying this process are not fully understood. The aim of this paper was to investigate the simultaneous removal of hydrophobic hexane with hydrophilic ethanol, resulting in the enhanced removal of hexane in the presence of ethanol. Investigations were performed in a peat-perlite packed biotrickling filter and the process performance was monitored using both gas chromatography and electronic nose techniques. The results indicate that the length as well as the efficiency of biofiltration during the start-up period depend on the feed composition, with higher efficiency obtained when hexane and ethanol were fed together from the process initiation. The experiments in the steady-state period present the biofilter performance when different ratios of hydrophilic to hydrophobic compounds were fed to the biofilter. The obtained results show the synergistic effects of the addition of a hydrophilic compound on the removal efficiency of hydrophobic hexane. The influence of the ratio of hydrophilic to hydrophobic compounds is discussed in terms of enhancing the mass transfer phenomena for hydrophobic volatile organic compounds.

Determination of Odor Intensity of Binary Gas Mixtures Using Perceptual Models and an Electronic Nose Combined with Fuzzy Logic

Measurement and monitoring of air quality in terms of odor nuisance is an important problem. From a practical point of view, it would be most valuable to directly link the odor intensity with the results of analytical air monitoring. Such a solution is offered by electronic noses, which thanks to the possibility of holistic analysis of the gas sample, allow estimation of the odor intensity of the gas mixture. The biggest problem is the occurrence of odor interactions between the mixture components. For this reason, methods that can take into account the interaction between components of the mixture are used to analyze data from the e-nose. In the presented study, the fuzzy logic algorithm was proposed for determination of odor intensity of binary mixtures of eight odorants: n-Hexane, cyclohexane, toluene, o-xylene, trimethylamine, triethylamine, α-pinene, and β-pinene. The proposed algorithm was compared with four theoretical perceptual models: Euclidean additivity, vectorial additivity, U model, and UPL model.

Mixed Natural Gas Online Recognition Device Based on a Neural Network Algorithm Implemented by a FPGA

It is a daunting challenge to measure the concentration of each component in natural gas, because different components in mixed gas have cross-sensitivity for a single sensor. We have developed a mixed gas identification device based on a neural network algorithm, which can be used for the online detection of natural gas. The neural network technology is used to eliminate the cross-sensitivity of mixed gases to each sensor, in order to accurately recognize the concentrations of methane, ethane and propane, respectively. The neural network algorithm is implemented by a Field-Programmable Gate Array (FPGA) in the device, which has the advantages of small size and fast response. FPGAs take advantage of parallel computing and greatly speed up the computational process of neural networks. Within the range of 0–100% of methane, the test error for methane and heavy alkanes such as ethane and propane is less than 0.5%, and the response speed is several seconds.

Monitoring of Cell Concentration during Saccharomyces cerevisiae Culture by a Color Sensor: Optimization of Feature Sensor Using ACO

The odor information produced in Saccharomyces cerevisiae culture is one of the important characteristics of yeast growth status. This work innovatively presents the quantitative monitoring of cell concentration during the yeast culture process using a homemade color sensor. First, a color sensor array, which could visually represent the odor changes produced during the yeast culture process, was developed using eleven porphyrins and one pH indicator. Second, odor information of the culture substrate was obtained during the process using the homemade color sensor. Next, color components, which came from different color sensitive spots, were extracted first and then optimized using the ant colony optimization (ACO) algorithm. Finally, the back propagation neural network (BPNN) model was developed using the optimized feature color components for quantitative monitoring of cell concentration. Results demonstrated that BPNN models, which were developed using two color components from FTPPFeCl (component B) and MTPPTE (component B), can obtain better results on the basis of both the comprehensive consideration of the model performance and the economic benefit. In the validation set, the average of determination coefficient R P 2 was 0.8837 and the variance was 0.0725, while the average of root mean square error of prediction (RMSEP) was 1.0033 and the variance was 0.1452. The overall results sufficiently demonstrate that the optimized sensor array can satisfy the monitoring accuracy and stability of the cell concentration in the process of yeast culture.

A New Method of Mixed Gas Identification Based on a Convolutional Neural Network for Time Series Classification

This paper proposes a new method of mixed gas identification based on a convolutional neural network for time series classification. In view of the superiority of convolutional neural networks in the field of computer vision, we applied the concept to the classification of five mixed gas time series data collected by an array of eight MOX gas sensors. Existing convolutional neural networks are mostly used for processing visual data, and are rarely used in gas data classification and have great limitations. Therefore, the idea of mapping time series data into an analogous-image matrix data is proposed. Then, five kinds of convolutional neural networks-VGG-16, VGG-19, ResNet18, ResNet34 and ResNet50-were used to classify and compare five kinds of mixed gases. By adjusting the parameters of the convolutional neural networks, the final gas recognition rate is 96.67%. The experimental results show that the method can classify the gas data quickly and effectively, and effectively combine the gas time series data with classical convolutional neural networks, which provides a new idea for the identification of mixed gases.

Functional Nanoparticles-Coated Nanomechanical Sensor Arrays for Machine Learning-Based Quantitative Odor Analysis

A sensing signal obtained by measuring an odor usually contains varied information that reflects an origin of the odor itself, while an effective approach is required to reasonably analyze informative data to derive the desired information. Herein, we demonstrate that quantitative odor analysis was achieved through systematic material design-based nanomechanical sensing combined with machine learning. A ternary mixture consisting of water, ethanol, and methanol was selected as a model system where a target molecule coexists with structurally similar species in a humidified condition. To predict the concentration of each species in the system via the data-driven approach, six types of nanoparticles functionalized with hydroxyl, aminopropyl, phenyl, and/or octadecyl groups were synthesized as a receptor coating of a nanomechanical sensor. Then, a machine learning model based on Gaussian process regression was trained with sensing data sets obtained from the samples with diverse concentrations. As a result, the octadecyl-modified nanoparticles enhanced prediction accuracy for water while the use of both octadecyl and aminopropyl groups was indicated to be a key for a better prediction accuracy for ethanol and methanol. As the prediction accuracy for ethanol and methanol was improved by introducing two additional nanoparticles with finely controlled octadecyl and aminopropyl amount, the feedback obtained by the present machine learning was effectively utilized to optimize material design for better performance. We demonstrate through this study that various information which was extracted from plenty of experimental data sets was successfully combined with our knowledge to produce wisdom for addressing a critical issue in gas phase sensing.

Wound odor: current methods of treatment and need for objective measures

Chronic wounds are an enormous burden to society, costing billions of dollars annually in the USA alone. Despite the extensive research into methods to heal chronic wounds, many remain unhealed for months to years. There is a need to focus on patient reported outcomes to improve quality of life in patients with non-healing wounds. Wound odor has a significant impact on patient quality of life; however, relatively little information is available on the management of wound odor. We review the current data available on wound odor and discuss the need for standardized objective measures of odor to improve research quality. An independent search of the PubMed and Embase databases was conducted using combinations of the following words or phrases: "wounds," "chronic wounds," "diabetic ulcers," "venous leg ulcers (VLUs)," "malignant ulcers," "odor," "odour," "smell," "malodor," "artificial olfaction," "electronic nose," and "e-nose." Article references were also searched for significance. There are few overall studies on wound odor, and fewer randomized controlled trials. Current trials on odor have consistent weaknesses such as subjective measures and poor methodology. No single odor treatment modality has been demonstrated to be widely effective for wound odor or superior to other methods. Future research should incorporate objective measures of odor such as electronic noses into clinical trials.

Online Sensor Drift Compensation for E-Nose Systems Using Domain Adaptation and Extreme Learning Machine

Sensor drift is a common issue in E-Nose systems and various drift compensation methods have received fruitful results in recent years. Although the accuracy for recognizing diverse gases under drift conditions has been largely enhanced, few of these methods considered online processing scenarios. In this paper, we focus on building online drift compensation model by transforming two domain adaptation based methods into their online learning versions, which allow the recognition models to adapt to the changes of sensor responses in a time-efficient manner without losing the high accuracy. Experimental results using three different settings confirm that the proposed methods save large processing time when compared with their offline versions, and outperform other drift compensation methods in recognition accuracy.

Determination of Odour Interactions in Gaseous Mixtures Using Electronic Nose Methods with Artificial Neural Networks

This paper presents application of an electronic nose prototype comprised of eight sensors, five TGS-type sensors, two electrochemical sensors and one PID-type sensor, to identify odour interaction phenomenon in two-, three-, four- and five-component odorous mixtures. Typical chemical compounds, such as toluene, acetone, triethylamine, α-pinene and n-butanol, present near municipal landfills and sewage treatment plants were subjected to investigation. Evaluation of predicted odour intensity and hedonic tone was performed with selected artificial neural network structures with the activation functions tanh and Leaky rectified linear units (Leaky ReLUs) with the parameter a=0.03. Correctness of identification of odour interactions in the odorous mixtures was determined based on the results obtained with the electronic nose instrument and non-linear data analysis. This value (average) was at the level of 88% in the case of odour intensity, whereas the average was at the level of 74% in the case of hedonic tone. In both cases, correctness of identification depended on the number of components present in the odorous mixture.

Comparison of SVM, RF and ELM on an Electronic Nose for the Intelligent Evaluation of Paraffin Samples

Paraffin odor intensity is an important quality indicator when a paraffin inspection is performed. Currently, paraffin odor level assessment is mainly dependent on an artificial sensory evaluation. In this paper, we developed a paraffin odor analysis system to classify and grade four kinds of paraffin samples. The original feature set was optimized using Principal Component Analysis (PCA) and Partial Least Squares (PLS). Support Vector Machine (SVM), Random Forest (RF), and Extreme Learning Machine (ELM) were applied to three different feature data sets for classification and level assessment of paraffin. For classification, the model based on SVM, with an accuracy rate of 100%, was superior to that based on RF, with an accuracy rate of 98.33-100%, and ELM, with an accuracy rate of 98.01-100%. For level assessment, the R² related to the training set was above 0.97 and the R² related to the test set was above 0.87. Through comprehensive comparison, the generalization of the model based on ELM was superior to those based on SVM and RF. The scoring errors for the three models were 0.0016-0.3494, lower than the error of 0.5-1.0 measured by industry standard experts, meaning these methods have a higher prediction accuracy for scoring paraffin level.