Model R2 Research Articles

Fractional derivative Burgers models describing dynamic viscoelastic properties of asphalt binders

The Burgers model is ineffective in the representation of dynamic viscoelastic properties of asphalt binders; the existing fractional derivative Burgers (FDB I) model has limitations in its application and construction concept. Consequently, this study started by deriving the constitutive equations of the FDB I model and proposed a new application idea of Abel dashpot to construct the second fractional derivative Burgers (FDB II) model. Then, dynamic modulus formulas for the FDB II model were derived. On this basis, the frequency sweep tests of base and two polyphosphoric acid (PPA) modified asphalt binders were carried, and the dynamic viscoelastic properties of three different asphalt binders were described using three different models, and in-depth comparisons and analysis are carried out. The results show that both fractional derivative Burgers (FDB) models can simultaneously express the dynamic viscoelastic properties of asphalt binders with a set of parameters. Both FDB models perform better than traditional Burgers model in the representation of dynamic viscoelastic properties, and fitting results of master curves no longer exhibit oscillation properties like the traditional Burgers model. The R2 of FDB models could reach above 0.99, while R2 of traditional Burgers model was only 0.9349 at minimum. Among them, the FDB II model performs best. Five parameters of the FDB II model have relatively clear physical meanings. The new application concept of fractional dashpots was verified its theoretical feasibility and application advantages, which will provide a new idea for further application and development of fractional derivative constitutive models.

Development of machine learning models for estimating wheat biophysical variables using satellite-based vegetation indices

Monitoring wheat biophysical variables during the growing season is essential for achieving precision agriculture's important goals. This study estimated plant height, leaf development stage, leaf area index, dry matter, and plant nitrogen content in wheat from Landsat-7 images in North Iran using machine learning algorithms. Forty points were selected from ten wheat fields (270 ha), and the agronomic parameters were recorded during the 2011–2012 growing season. The machine learning models applied in this study were Artificial Neural Network (ANN), Support Vector Machine (SVM), and Deep Neural Network (DNN). Results suggest that the DNN model was the best for predicting all the parameters, except for the nitrogen percentage, in which SVM and DNN have the same accuracy. The DNN model R2 was 0.82, 0.95, 0.95, and 0.90 RMSE (Root Mean Square Error) 9.61, 0.46, 0.47 and 1.2 and MAE (Mean Absolute Error) 26.4, 0.34, 0.28, and 0.84 for estimating plant height, leaf area index, leaf development stage and dry matter biomass, respectively. The DNN and SVM algorithms can effectively predict plant nitrogen percentage with identical R2, RMSE, and MAE values of 0.83–0.84, 0.44–0.46, and 0.33–0.36, respectively. These results show that machine learning models have great potential to provide timely and accurate data for better wheat management under field conditions from satellite imagery.

Investigating on the Pavement Performance of Multi-Source Solid Wastes by Cement and Fly Ash.

In order to advance the utilization rate of multi-source solid wastes in the Ningxia region of China, 16 groups of pavement base mixtures were designed with cement and fly ash (FA) as binders, steel slag (SS), silicon manganese slag (SMS), and recycled crushed stone (RCS) as composite aggregates. The evolution laws of mechanical and frost resistance properties of the mixture were investigated by unconfined compressive strength (UCS), indirect tensile strength (ITS), freeze-thaw (FT), and ultrasonic detection tests. Then, the strength formation mechanisms were revealed by microscopic characterization technology. The mathematical models between UCS-ITS, UCS-ultrasonic amplitude, FT cycles-UCS damage, and frost resistance coefficient-relative dynamic elastic modulus Er were established. The results show that cement content and curing age exhibited a positive effect on the mechanical strength and frost resistance of the mixture. When the replacement rate of SS was 60%, the mechanical strength and frost resistance were preferable. The R2 of the strength relationship models constructed was greater than 0.9, indicating high fitting accuracy. With the extension of the curing age, the cementitious products such as C-S-H (hydrated calcium silicate) and AFt (ettringite) developed entirely, and they were interlocked and cemented with each other, resulting in the micro-morphology developed from the three-dimensional network structure to the dense system. The macroscopic behavior incarnated that the mechanical strength and frost resistance of the mixture were significantly enhanced.

Association between Oxygen Reserve index and arterial partial pressure of oxygen during one-lung ventilation: a retrospective cohort study.

We aimed to investigate the association between the Oxygen Reserve index (ORi) and arterial partial pressure of oxygen (PaO2) during one-lung ventilation in patients who underwent non-cardiac thoracic surgery requiring one-lung ventilation. This retrospective study assessed the eligibility of 578 adult patients who underwent elective non-cardiac thoracic surgery requiring one-lung ventilation at a tertiary hospital, and their electronic medical records were reviewed. The ORi monitor was used in all patients during anesthesia, and arterial blood gas analysis was routinely performed 15 min after the initiation of one-lung ventilation. The primary endpoint was the association between ORi and PaO2 which were measured simultaneously during one-lung ventilation. We also investigated the risk factors for PaO2 less than 150 mmHg during one-lung ventilation. Total of 554 patient were included in the analysis. The ORi value measured 15 min after the start of one-lung ventilation was significantly associated with PaO2 in the linear regression model (r2 = 0.5752, P < 0.001), and 0.27 of the ORi value could distinguish PaO2 ≥ 150 mmHg (sensitivity 0.909, specificity 0.932). Risk factors for PaO2 < 150 mmHg during one-lung ventilation included a lower ORi, older age, higher body mass index, left-sided one-lung ventilation, and lower hemoglobin concentrations. This study suggested that ORi could provide useful information on arterial oxygenation even during one-lung ventilation for non-cardiac thoracic surgery.

Ocean ambient noise field modelling and the optimized noise term

The objective of this thesis is to determine the frequency and wind-wave forcing dependent effective sea surface noise source level per unit area extracted from the hourly minimum sound power levels of six month-long acoustic recordings. The effect of the propagation environment is accounted for using Bellhop. The simulated environment is configured using climatological sound velocity profiles to capture seasonal effects and bottom sound speed estimates made from seabed sediment maps. Hourly meteorological data were extracted from ERA5 providing relevant wind and wave parameters from which noise levels may be predicted. A weighted composite model consisting of neutral wind and significant wave height leveraging the two-term exponential regression function proved to maximize model R2. Received level data originating from 16 hydrophone stations in the North Atlantic and Labrador Sea were combined with the Bellhop TL simulations in order to produce estimates of the effective noise source level per unit area (NSL/A) for changing surface environmental conditions and inter-compared. Hourly minimum sound power level derived model-data comparisons using horizontal wind speed magnitude 10 m above sea level expressed a decrease in NSL/A estimates versus Kewley (1990) by 10 to 15 dB from 1 to 3 kHz.

Desulphurization of Simulated Oil Using SAPO-11 with CNT's as Adsorbent: A Kinetic Study

In this study, sulfur was removed from imitation oil using oxidative desulfurization process. Silicoaluminophosphate (SAPO-11) was prepared using the hydrothermal method with a concentration of carbon nanotubes (CNT) of 0% and 7.5% at 190 °C crystallization temperature. The final molar composition of the as-prepared SAPO-11 was Al2O3: 0.93P2O5: 0.414SiO2. 4% MO/SAPO-11 was prepared using impregnation methods. The produced SAPO-11 was described using X-ray diffraction (XRD) and Brunauer-Emmet-Teller (N2 adsorption–desorption isotherms). It was found that the addition of CNT increased the crystallinity of SAPO-11. The results showed that the surface area of SAPO-11 containing 7.5% CNT was 179.54 m2/g, and the pore volume was 0.317 cm3/g. However, the surface area of SAPO-11 containing 0% CNT was 125.311 m2/g, and pore volume was 0.275 cm3/g, while nanoparticles with an average particle diameter of 24.8 nm were obtained. Then, the prepared SAPO-11 was used in the oxidative desulfurization process. The oxidative desulfurization was studied using several factors affecting desulfurization efficiency, such as time (40, 60, 80, 100, and 120) min, amount of MO/SAPO-11 (0.3, 0.4, 0.5, 0.6, and 0.7) g/100 ml of simulated oil (100 ppm of dibenzothiophene), the amount of hydrogen peroxide (4ml) oxidizer/100 ml of simulated oil, and the temperature ranges from (40, 50, 60, 70, and 80 °C). The results showed that an increase in MO/SAPO-11 led to an increase in desulfurization. The best removal percentage for sulfur content was 92.79%, obtained at 70 °C and 0.6 g of MO/SAPO-11 containing 7.5% CNT, and the removal was 82.34% at 0% CNT and the same other conditions. While the equilibrium was achieved after 100 min. The results revealed that Freundlich's model described the adsorption of sulfur compounds better than Langmuir's, where the R2 of the Freundlich model was 0.9979 and the R2 of the Langmuir model was 0.9554.

Combining Vis-NIR and NIR Spectral Imaging Techniques with Data Fusion for Rapid and Nondestructive Multi-Quality Detection of Cherry Tomatoes.

Firmness, soluble solid content (SSC) and titratable acidity (TA) are characteristic substances for evaluating the quality of cherry tomatoes. In this paper, a hyper spectral imaging (HSI) system using visible/near-infrared (Vis-NIR) and near-infrared (NIR) was proposed to detect the key qualities of cherry tomatoes. The effects of individual spectral information and fused spectral information in the detection of different qualities were compared for firmness, SSC and TA of cherry tomatoes. Data layer fusion combined with multiple machine learning methods including principal component regression (PCR), partial least squares regression (PLSR), support vector regression (SVR) and back propagation neural network (BP) is used for model training. The results show that for firmness, SSC and TA, the determination coefficient R2 of the multi-quality prediction model established by Vis-NIR spectra is higher than that of NIR spectra. The R2 of the best model obtained by SSC and TA fusion band is greater than 0.9, and that of the best model obtained by the firmness fusion band is greater than 0.85. It is better to use the spectral bands after information fusion for nondestructive quality detection of cherry tomatoes. This study shows that hyperspectral imaging technology can be used for the nondestructive detection of multiple qualities of cherry tomatoes, and the method based on the fusion of two spectra has a better prediction effect for the rapid detection of multiple qualities of cherry tomatoes compared with a single spectrum. This study can provide certain technical support for the rapid nondestructive detection of multiple qualities in other melons and fruits.

Prediction of pasture yield using machine learning-based optical sensing: a systematic review

AbstractAccurate and reliable predictions of biomass yield are important for decision-making in pasture management including fertilization, pest control, irrigation, grazing, and mowing. The possibilities for monitoring pasture growth and developing prediction models have greatly been expanded by advances in machine learning (ML) using optical sensing data. To facilitate the development of prediction models, an understanding of how ML techniques affect performance is needed. Therefore, this review examines the adoption of ML-based optical sensing for predicting the biomass yield of managed grasslands. We carried out a systematic search for English-language journal articles published between 2015-01-01 and 2022-10-26. Three coders screened 593 unique records of which 91 were forwarded to the full-text assessment. Forty-three studies were eligible for inclusion. We determined the adoption of techniques for collecting input data, preprocessing, and training prediction models, and evaluating their performance. The results show (1) a broad array of vegetation indices and spectral bands obtained from various optical sensors, (2) an emphasis focus on feature selection to cope with high-dimensional sensor data, (3) a low reporting rate of unitless performance metrics other than R2, (4) higher variability of R2 for models trained on sensor data of larger distance from the pasture sward, and (5) the need for greater comparability of study designs and results. We submit recommendations for future research and enhanced reporting that can help reduce barriers to the integration of evidence from studies.

Integrated data mining for prediction of specific capacitance of porous carbon materials for flexible energy storage devices

Porous carbon materials have gained wide attention owing to their remarkable electrical conductivity and large surface area. Predicting the specific capacitance based on these materials is a crucial step towards designing and manufacturing flexible energy storage devices. To achieve this objective, we employed integrated data mining to construct a database comprising 18 input variables (such as precursor material, activation type, reference electrode, electrolyte, annealing temperature, annealing time, activation temperature, ramp rate of activation temperature, activation time, specific surface area of the pore, total volume of the pore, mesoporous volume, micropore volume, current density, activator material ratio, mass fraction, average pore size, nitrogen content) and one output variable (specific capacitance). We selected characteristic variables to build a porous carbon electrode material database and imported it into a layer fusion model and single-model such as XGBoost, LightGBM, and linear regression to calculate the influence ranking of characteristic variables. The layer fusion model was compared with the single-model by importing a blind dataset into the training, and the prediction correlation coefficient R2 of the layer fusion model was found to be 0.981, indicating superior performance compared to the single-model. This layer fusion model enables simulation predictions for the selection and preparation of porous carbon electrode materials, guides experimental design, optimizes experimental schemes, and targets experimental parameters by referencing the influence of characteristic variables to obtain optimal experimental results.

Accelerated Design for High-Entropy Alloys Based on Machine Learning and Multiobjective Optimization.

High-entropy alloys (HEAs) with high hardness and high ductility can be considered as candidates for wear-resistant applications. However, designing novel HEAs with multiple desired properties using traditional alloy design methods remains challenging due to the enormous composition space. In this work, we proposed a machine-learning-based framework to design HEAs with high Vickers hardness (H) and high compressive fracture strain (D). Initially, we constructed data sets containing 172,467 data with 161 features for D and H, respectively. Four-step feature selection was performed, with the selection of 12 and 8 features for the D and H prediction models based on the optimal algorithms of the support vector machine (SVR) and light gradient boosting machine (LightGBM), respectively. The R2 of the well-trained models reached 0.76 and 0.90 for the 10-fold cross validation. Nondominated sorting genetic algorithm version II (NSGA-II) and virtual screening were employed to search for the optimal alloying compositions, and four recommended candidates were synthesized to validate our methods. Notably, the D of three candidates have shown significant improvements compared to the samples with similar H in the original data sets, with increases of 135.8, 282.4, and 194.1% respectively. Analyzing the candidates, we have recommended suitable atomic percentage ranges for elements such as Al (2-14.8 at %), Nb (4-25 at %), and Mo (3-9.9 at %) in order to design HEAs with high hardness and ductility.

Early factors associated with continuous positive airway pressure failure in moderate and late preterm infants

To determine the early factors associated with continuous positive airway pressure (CPAP) failure in moderate-to-late preterm infants (32 + 0/7 to 36 + 6/7 weeks’ gestation) from the NEOBS cohort study. The NEOBS study was a multi-center, prospective, observational study in 46 neonatal intensive care units in France, which included preterm and late preterm infants with early neonatal respiratory distress. This analysis included a subset of the NEOBS population who had respiratory distress and required ventilatory support with CPAP within the first 24 h of life. CPAP failure was defined as the need for tracheal intubation within 72 h of CPAP initiation. Maternal and neonatal clinical parameters in the delivery room and clinical data at 3 h of life were analyzed. CPAP failure occurred in 45/375 infants (12%), and compared with infants with CPAP success, they were mostly singletons (82.2% vs. 62.1%; p < 0.01), had a lower Apgar score at 10 min of life (9.1 ± 1.3 vs. 9.6 ± 0.8; p = 0.02), and required a higher fraction of inspired oxygen (FiO2; 34.4 ± 15.9% vs. 22.8 ± 4.1%; p < 0.0001) and a higher FiO2*positive end-expiratory pressure (PEEP) (1.8 ± 0.9 vs. 1.1 ± 0.3; p < 0.0001) at 3 h. FiO2 value of 0.23 (R2 = 0.73) and FiO2*PEEP of 1.50 (R2 = 0.75) best predicted CPAP failure. The risk of respiratory distress and early CPAP failure decreased 0.7 times per 1-week increase in gestational age and increased 1.7 times with every one-point decrease in Apgar score at 10 min and 19 times with FiO2*PEEP > 1.50 (vs. ≤ 1.50) at 3 h (R2 of the overall model = 0.83). Conclusion: In moderate-to-late preterm infants, the combination of singleton pregnancy, lower Apgar score at 10 min, and FiO2*PEEP > 1.50 at 3 h can predict early CPAP failure with increased accuracy.What is Known:•Respiratory distress syndrome (RSD) represents an unmet medical need in moderate-to-late preterm births and is commonly treated with continuous positive airway pressure (CPAP) to reduce mortality and the need for additional ventilatory support.• Optimal management of RSD is yet to be established, with several studies suggesting that identification of predictive factors for CPAP failure can aid in the prompt treatment of infants likely to experience this failure.What is New:•Secondary analysis of the observational NEOBS study indicated that oxygen requirements during CPAP therapy, especially the product of fraction of inspired oxygen (FiO2) and positive end-expiratory pressure (PEEP), are important factors associated with early CPAP failure in moderate-to-late term preterm infants.•The combination of a singleton pregnancy, low Apgar score at 10 minutes, and high FiO2*PEEP at 3 hours can predict early CPAP failure with increased accuracy, highlighting important areas for future research into the prevention of CPAP failure.

Development and temporal validation of clinical prediction models for 1-year disability and pain after lumbar decompressive surgery. The Norwich Lumbar Surgery Predictor (development version).

To identify clinical predictors and build prediction models for 1-year patient-reported outcomes measures (PROMs) after lumbar decompressive surgery for disc herniation or spinal stenosis. The study included 1835 cases, with or without additional single-level fusion, from a single centre from 2008 through 2020. General linear models imputed with 37 clinical variables identified 18 significant 1-year PROM predictors for retention in development models. Interaction of surgical indication with each predictor was tested. Temporal validation was conducted at the same centre on cases through 2021. R2 was used to measure goodness-of-fit, and area under curve (AUC) used to measure classification to a satisfactory symptom state (Oswestry Disability Index (ODI) ≤ 22; back or leg pain ≤ 30 out of 100). A total 1228 (67%) had complete data for inclusion in model development. Predictors of ODI were baseline PROMs (ODI, back pain, leg pain), work status, condition duration, previous lumbar operation, multiple-joint osteoarthritis, female, diabetes, current smoker, rheumatic disorder, lower limb arthroplasty, mobility aided, provider status, facet cyst, scoliosis, and age, with BMI significantly associated with stenosis. Temporal validation (n = 188) found the ODI model R2 was 0.29 (95% confidence intervals (CI) 0.18-0.40) and AUC was 0.74 (95% CI 0.67-0.81). Back and leg pain models had lower R2 (0.12-0.14) and AUC (0.68-0.69) values. Important PROM predictors are baseline PROMs, specific co-morbidities, work status, condition duration, previous lumbar operation, female, and smoking status. The ODI model predicted the likelihood of achieving a satisfactory state of both disability and pain.

Machine Learning Aided Prediction of Glass-Forming Ability of Metallic Glass

The prediction of the glass-forming ability (GFA) of metallic glasses (MGs) can accelerate the efficiency of their development. In this paper, a dataset was constructed using experimental data collected from the literature and books, and a machine learning-based predictive model was established to predict the GFA. Firstly, a classification model based on the size of the critical diameter (Dmax) was established to determine whether an alloy system could form a glass state, with an accuracy rating of 0.98. Then, regression models were established to predict the crystallization temperature (Tx), glass transition temperature (Tg), and liquidus temperature (Tl) of MGs. The R2 of the prediction model obtained in the test set was greater than 0.89, which showed that the model had good prediction accuracy. The key features used by the regression models were analyzed using variance, correlation, embedding, recursive, and exhaustive methods to select the most important features. Furthermore, to improve the interpretability of the prediction model, feature importance, partial dependence plot (PDP), and individual conditional expectation (ICE) methods were used for visualization analysis, demonstrating how features affect the target variables. Finally, taking Zr-Cu-Ni-Al system MGs as an example, a prediction model was established using a genetic algorithm to optimize the alloy composition for high GFA in the compositional space, achieving the optimal design of alloy composition.

Prediction of glaciated area fraction over the Sikkim Himalayan Region, India: a comparative study using response surface method, random forest, and artificial neural network.

Glacier area fraction at high altitude mountains is a serious worry in today's time triggered by climate change. The current information on this natural resource is very important for the survival of humanity as it affects the water, food, and energy security of people dependent on it. Due to its problematic accessibility and tough environmental condition, ground monitoring is quite challenging. This study investigates the impact of environmental parameters and pollutants on glacier area fraction over the Eastern Himalaya region and its prediction through random forest (RF), multilayer perceptron (MLP), radial basis function analysis (RBFN), and response surface methodology (RSM) models. The data are obtained from the Goddard Earth Sciences Data and Information Services Center (GES DISC), NASA's data archive portal ( https://giovanni.gsfc.nasa.gov ). The collinearity of independent variables reveals that all selected input parameters are highly correlated with R2 value > 0.9. The RSM and RF model provided valuable insight of the predictor's significance in addition to their capability to predict the response. The model performance was evaluated in terms of R2 value and the error matrices. The model's R2 value was found to be 0.843, 0.839, 0.838, and 0.743 for MLP, RBFN, RF, and RSM respectively. Although, the neural network model R2 values are the highest, but the most reliable and suitable model is RF as the error matrices for this model are much lower than others. This study encourages the investigation of the hybridization of these models for more accurate prediction.

Study on Performance Optimization of Water-Rich Grouting Materials Based on Response Surface Methodology

The quality of borehole sealing is a key factor affecting the efficiency of gas production. A new water-rich grouting material (RW) with composite coagulant and other additives was prepared in this study to overcome the disadvantages of long setting time and low stone rate of traditional cement materials. When the coagulants A is 4 g and coagulants B is 2 g, the setting time of RW material was reduced by 60.85% and 50.62%, which significantly shortened the setting time of the RW material, respectively. Based on the orthogonal method, 29 groups of comparative experiments were designed to investigate the interaction mechanism between different additives on the performance index of RW, including setting time, water secretion rate, and compressive strength. Quadratic regression equations were fitted using the response surface method. All the correlation coefficients R2 of each response model were greater than 0.97, R2 and R2adj were less than 0.2 through variance analysis, indicating a high correlation between the actual and prediction results. The water–cement ratio had the most significant effect among all factors on setting time, water secretion rate, and compressive strength of the RW material. The scanning electron microscope (SEM) was used to compared the micromorphological characteristics of RW and conventional Portland cement material (PC). The results showed that the hydration products of RW were mostly smack ettringite, calcium silicate hydrate gel, and calcium hydroxide, which interweaved with each other to form a network structure that was denser than the PC material. Furthermore, the interface bonding degree between RW and injected coal was tighter than that of PC, without obvious cracks at the slurry–coal interface. The results indicate that the addition of composite coagulant can significantly accelerate the hydration process of RW material and also enhance the interface strength of injected coal, which is conducive to improving the grouting quality and sealing effect of the extraction borehole.

Simulating fate and transport of non-steroidal anti-inflammatory drugs using Root Zone Water Quality Model 2.

The fate and transport of non-steroidal anti-inflammatory drugs (NSAIDs) in soil are determined by various processes, and the complexity of the system lends itself to the use of computer simulation models to help understand it. This study demonstrated the first attempt to use empirical data from lab incubation and field studies to parameterize and test a process-based agricultural systems model, Root Zone Water Quality Model 2 (RZWQM2), to simulate the fate and transport of naproxen (NPX), ibuprofen (IBF), and ketoprofen (KTF) in field-based lysimeters amended with alkaline-treated biosolids (ATBs). The model calibrated for the soil-water balance module and contaminant transport module was used to predict water seepage through the soil profile in 2017 and 2018 within a 15% error of the field measured data, with model performance statistics such as Nash-Sutcliffe model efficiency (NSE) and R2 all greater than 0.70. The overall predicted percent recovery of initial spiked NSAIDs in both soil and water samples, after further calibration of the contaminant transport module, was within the same order of magnitude as the measured data. The model underestimated the percent recovery of initial spiked NSAIDs at the 30- to 55-cm soil depth for all treatments on day 3. The calibrated soil subsurface aerobic half-lives of NPX and IBF were found to be considerably lower than their laboratory-measured half-lives obtained from the incubation study. The overall performance of RZWQM2 in simulating the soil hydrology and behavior of NSAIDs in soil profiles receiving various rates of ATB amendments was satisfactory.

Combining machine learning with multi-physics modelling for multi-objective optimisation and techno-economic analysis of electrochemical CO2 reduction process

As a carbon capture and utilization (CCU) technology, gas diffusion electrode (GDE) based electrochemical CO2 reduction reaction (eCO2RR) can convert CO2 to valuable products, such as formate and CO. However, the electrode parameters and operational conditions need to be studied and optimised to enhance the performance and reduce the net cost of the eCO2RR process before its industrial application. In this work, a machine learning algorithm, i.e., extended adaptive hybrid functions (E-AHF) is combined with a multi-physics model for the data-driven three-objective optimisation and techno-economic analysis of the GDE-based eCO2RR process. The effects of eight design variables on the product yield (PY), CO2 conversion (CR) and specific electrical energy consumption (SEEC) of the process are analysed. The results show that the R2 of the E-AHF model for the prediction of PY, CR and SEEC are all higher than 0.96, indicating the high accuracy of the developed machine learning algorithm for the prediction of the eCO2RR process. The process performance experiences a notable improvement after optimisation and is affected by a combination of eight variables, amongst which the electrolyte concentration having the most significant impact on PY and CR. The optimal trade-off single-pass PY, CR and SEEC are 3.25×10−9 kg s−1, 0.663% and 9.95 kWh kg−1 based on flow channels with 1 cm in length, respectively. The SEEC is reduced by nearly half and PY and CR are improved more than two times after optimisation. The production cost of the GDE-based eCO2RR process was approximately $378 t−1product (CO and formate), much lower than that of traditional CO2 utilisation factories ($835 t−1product). The electricity cost accounted for more than 80% of the total cost, amounting to $318 t−1, indicating that cheaper and cleaner electricity sources would further reduce the production cost of the process, which is the key to the economics of this technology.

Machine Learning Analysis of Raman Spectra To Quantify the Organic Constituents in Complex Organic-Mineral Mixtures.

Important decisions in local agricultural policy and practice often hinge on the soil's chemical composition. Raman spectroscopy offers a rapid noninvasive means to quantify the constituents of complex organic systems. But the application of Raman spectroscopy to soils presents a multifaceted challenge due to organic/mineral compositional complexity and spectral interference arising from overwhelming fluorescence. The present work compares methodologies with the capacity to help overcome common obstacles that arise in the analysis of soils. We created conditions representative of these challenges by combining varying proportions of six amino acids commonly found in soils with fluorescent bentonite clay and coarse mineral components. Referring to an extensive data set of Raman spectra, we compare the performance of the convolutional neural network (CNN) and partial least-squares regression (PLSR) multivariate models for amino acid composition. Strategies employing volume-averaged spectral sampling and data preprocessing algorithms improve the predictive power of these models. Our average test R2 for PLSR models exceeds 0.89 and approaches 0.98, depending on the complexity of the matrix, whereas CNN yields an R2 range from 0.91 to 0.97, demonstrating that classic PLSR and CNN perform comparably, except in cases where the signal-to-noise ratio of the organic component is very low, whereupon CNN models outperform. Artificially isolating two of the most prevalent obstacles in evaluating the Raman spectra of soils, we have characterized the effect of each obstacle on the performance of machine learning models in the absence of other complexities. These results highlight important considerations and modeling strategies necessary to improve the Raman analysis of organic compounds in complex mixtures in the presence of mineral spectral components and significant fluorescence.

Shared E-Scooter User Characteristics and Usage Patterns across Four U.S. Cities

The rapid growth of shared e-scooters around the world creates an interest and a need to understand who uses shared e-scooters, trip patterns, and modal shift impacts. To shed light on these questions, this study conducted an online survey ( N = 1498) to collect information on shared e-scooter use in four U.S. cities: Birmingham, AL, Washington D.C., Los Angeles, CA, and Miami, FL. The paper provides a comprehensive view of shared e-scooter use in different cities and reveals the similarities and differences in e-scooter users’ socioeconomic characteristics. Furthermore, the results of a binary logistic model show the impact of socioeconomic and travel modal choice factors on predicting shared e-scooter usage. The findings suggest that male, younger, and more affluent users who live in smaller households owning fewer vehicles are more likely to use shared e-scooter services. The relationships between income level, race/ethnicity, or e-scooter ownership and the usage of shared e-scooters are not statistically significant. The McFadden’s R2 of the binary logistic regression model indicates an excellent fit. The results also show that compared with non-users, shared e-scooter users tend to assess travel cost as more important and travel safety as less important in their mode choice. The findings from this work can help city planners, policymakers, and other micromobility stakeholders in their efforts to promote the adoption of shared e-scooters and improve on deployment practices of shared e-scooters at their locations.

Experimental study and small-sample-regression prediction on effect of carbon-based reinforcements on thermal conductivity of composite mortar

To reveal the impact of various sizes of inert reinforcement on the pore structure, hydration products, and microscopic morphology of mortar. To explore its effect on mortar's thermal conductivity. Investigated by microscopic tests and small sample regression modeling. It is found that nanoscale materials have a nanocore effect, which can effectively enhance hydration, but the filling and bridging of larger pores is inadequate, leading to lower thermal conductivity in the early stage. The release of their properties is limited by the dispersion technique. Micron-reinforcing-phase, easily dispersed and with adsorption, can effectively fill capillary pores, and adsorb the hydration products to advance the hydration process. Milli-reinforcing-phase, flocculent fiber water absorption is obvious, resulting in the reduction of the real water-cement ratio, hydration process is delayed. With age growth and water precipitation, the fiber surface produces nucleation sites, which will accelerate the hydration rate. For thermal conductivity, porosity is the main influence factor in the early curing stage; when the porosity changes are small, the internal microstructure and hydration products will become the dominant factors. The sample set is expanded by feature enhancement and data synthesis to improve the generalization ability. The predicted model R2 can reach 0.75, demonstrating that a limited sample of experimental data can make reliable predictions.