Input Variables Research Articles

CO2 Emission Efficiency Measurement: Green Logistics Perspective

Purpose: This study aims to measure the carbon emission activities of Turkey, Azerbaijan, Kazakhstan, Kyrgyzstan and Uzbekistan and evaluate them from a green logistics perspective. Method: Within the scope of the research, efficiency analyses were conducted using the output-oriented Data Envelopment Analysis (DEA) constant returns to scale (CRS) model and the super efficiency CRS DEA model. The input and output variables used in the research were obtained from the World Bank website. Findings: In the research, the efficiency scores of the relevant countries were determined. It was determined that the efficiency scores of the countries were generally above 50%. The ranking of the efficient decision-making units among themselves was carried out with the super efficiency CRS DEA model. Some potential improvement suggestions were presented for the decision-making units that were not efficient. Originality: In order to leave a livable world to future generations, green energy production should be supported and Carbon Dioxide (CO2) emissions should be kept under control. Therefore, the efficiency assessment of countries' CO2 emissions is of vital importance. This study has an original feature because the CO2 emission activities of the Turkish Republics were carried out using the super efficiency CRS DEA model. This study can provide guidance to those who will conduct research on this subject and to country leaders.

Powertrain Design and Modeling for a Fuel Cell Hybrid ElectricVehicle

The objective of this study was to develop a Fuel Cell Hybrid Electric Vehicle (FCHEV) powertrain with the aim of enhancing battery usage autonomy. The vehicle, which participated in the Eco-Marathon competition as a prototype, incorporates batteries, a Direct Current (DC) electric motor, and a Proton Exchange Membrane (PEM) fuel cell. The design permits the operation of the fuel cell to be conducted in a more efficacious and fuel-efficient manner. The study employs the MATLAB-Advisor software to construct powertrain models that are then validated in laboratory settings. These models are subsequently compared with the performance of the actual FCHEV prototype and adapted for use in automotive applications. The FCHEV power model calculates instantaneous energy consumption using input variables, such as vehicle speed, acceleration, and road gradient. Furthermore, Real Cycle drive was carried out to improve the trade-off between energy consumption, fuel cells, battery State of Charge (SOC) dynamics, and battery power smoothness, while ensuring that all essential limitations were met. The addition of a fuel cell to an electric car model enhances its range by 250%, significantly improving its adoption and usage.

Application of Machine Learning Predicting Injuries in Traffic Accidents through the Application of Random Forest

Background: The objective of this work is to analyze and predict the harmfulness in traffic accidents. Method: Several Random Forest statistical models are created, in which the predictable variable (response/ output variable) is the harmfulness of the accident, while the input variables are the various characteristics of the accident. In addition, these generated models will allow estimating the influence or importance of each of the factors studied (input variables) concerning the harmfulness of road accidents so that it is possible to know in which aspects it is more profitable to work with the objective of reducing mortality from traffic accidents [1]. Results: In this regard, the predictive algorithm has an out-of-bag error of 26.55% and an overall accuracy of 74.1%. Meanwhile, the local accuracy of the mildly wounded class is 66.1% compared to 81.4% of the dead and severely wounded class, which, as mentioned, has higher prediction reliability. Conclusion: Finally, it is worth noting the enormous usefulness of the Random Forest machine learning technique, which provides very useful information for possible research or studies that may be carried out. In the specific case of this work, through the use of the R programming language, which in turn presents a wide range of freely accessible utilities and functions with which it may be interesting working, it has generated results of great value for this area of activity, important to society as road safety.

An active learning strategy for sequential designs of probabilistic systems with continuous and discrete input variables

This study proposes an active learning strategy for sequential experimental designs of probabilistic systems with mixed continuous and discrete input variables. The proposed method aims to sequentially explore local optima with high aleatory uncertainties along with the global optimum of the system. In complex probabilistic systems, the exploration of both of these regions is needed to build a more robust understanding across different conditions or scenarios. The proposed strategy employs a non-deterministic kriging (NDK) method, capable of handling mixed input variables, to account for aleatory uncertainties. A modified version of the expected improvement concept is used to identify the regions of interest for the sequential experiments. To avoid neglecting local optima with high aleatory uncertainties, subset simulations are used to identify these regions. The methodology was applied to a set of analytical examples and yielded accurate predictions for the optimum values and associated aleatory uncertainties.

Identifying bacterial fixation pathway of mediating soil carbon stock changes along tropical forest restoration

The mechanism mediating carbon accumulation changes along ropical forest restoration remains unclear. Here, we identified how functional bacteria, litter input, and abiotic variables control soil organic carbon stock changes along an age-chronosequence of tropical forest restoration. Over 51-yr recovery, significant increases in total organic carbon stocks (∼1.7 fold) were strongly associated with increases in copy number of carbon fixation bacterial genes (cbbL) (∼2.6 fold). The direct pathways of cbbL abundance, microbial and mineral-associated organic carbon explained 76 % of carbon stock variation. In contrast, litter carbon, soil water, and bulk density indirectly regulated carbon stocks through affecting cbbL abundance (68 %) and microbial carbon level (29 %). Furthermore, cbbL abundance had a higher contribution (71 %) to carbon fraction transformation than microbial carbon level (19 %). We suggest that tropical forest restoration controls carbon stocks primarily via direct bacterial fixation pathway mediated by litter carbon and physical soil variables. Our results are helpful to further understand the mechanism of tropical forest restoration regulating carbon transformation and accumulation.

Efficient proxy for time-lapse seismic forward modeling using a U-net encoder-decoder approach

The time-lapse seismic (4D seismic) forward modeling provides crucial data for calibrating reservoir models through different data assimilation algorithms. Unfortunately, the traditional 4D seismic forward-modeling methodology is time-expensive and entails significant computational resource consumption. To address these drawbacks, in this work, our goal is to develop a proxy model for the 4D seismic forward modeling using a class of machine learning algorithm named U-Net encoder–decoder. We applied the developed proxy model to a benchmark carbonate reservoir using an ensemble of reservoir simulation models from UNISIM IV dataset (a synthetic benchmark based on real data of a Brazilian pre-salt field). Moreover, we aim to introduce seminal strategies for interpreting the proposed proxy model operation, its outputs, and possible correlations between input and output variables. To achieve this, we trained and tested two versions of U-net-based models and applied methods for explainable artificial intelligence, such as Gradient-weighted Class Activation Mapping (Grad-CAM) and Forward Feature Selection. The experiments showed good results when applied to the test dataset. The correlation coefficient (R2) values were in the range of 0.7 to 0.9, showing the efficiency of the proxy model to replace the 4D seismic forward modeling. Through qualitative analysis, it was possible to identify which input properties and regions of the reservoir are more relevant for the model’s inference. These results are a step toward robust, explainable machine learning-based proxy forward modeling.

Enhancing undrained shear strength prediction: a robust hybrid machine learning approach with naïve Bayes modeling

In geotechnical engineering, it is crucial to make sure that the undrained shear strength (USS) of soft, sensitive clays is accurately assessed. The accuracy in forecasting USS is pivotal for ensuring the structural integrity and stability of foundations and earthworks. Addressing this concern, advanced data-driven NB techniques are utilized to disclose the complex interactions of USS with basic soil parameters. This paper presents a novel methodology for the USS prediction in soft clays using machine learning techniques, and particularly it highlights the attention on the following five important input variables: pre-consolidation stress (PS), vertical effective stress (VES), liquid limit (LL), plastic limit (PL), and natural water content (W). These are selected in view of their well-understood impact on the USS. This study reports an innovative effort to use SHO and AOSM for the model's hyperparameter tuning, reducing heuristic methods and computationally expensive brute-force searches. This will provide a neat methodology for improving accuracy in USS predictions and maintaining the optimality of the model performance. The results, therefore, provide geotechnical engineers and researchers with considerable benefits. They give a sound basis that is data-driven for the assessment of USS in soft sensitive clays and advance the safety and stability of civil engineering projects.

Prediction of ketosis using radial basis function neural network in dairy cattle farming.

The purpose of the paper was to apply an Artificial Neural Networks with Radial Basis Function to develop an application model for diagnosing a subclinical ketosis type I and II in dairy cattle. While building the neural network model, applied methodology was compatible to the procedures used in Data Mining processes.The data set was created based on the composition of milk samples of 1520 Polish Holstein-Friesian cows. The milk samples were collected during test-day milkings and made available by Polish Federation of Cattle Breeders and Milk Producers. The milk composition parameters were used as the input variables for RBF network models. The value of the output variable was determined based on the content of β-hydroxybutyric acid in blood of cows. In the next stage of the work, the qualities of the pre-selected models were compared and the best ones were chosen. The sensitivity and specificity as well as the size of the AUC (Area Under the Curve) under the ROC (Receiver Operating Characteristic) were taken as the main criteria for network models evaluation. The model characterized by sensitivity of 0.86, specificity of 0.71 and AUC of 0.89 was selected for ketosis type I. The optimal for ketosis type II showed the sensitivity and specificity 0.81 and 0.75, respectively, and the size of AUC above 0.85. Chosen models were recorded using the predictive modelling markup language (PMML) for data mining models to be shared and used between the different applications.

Sediment flow connectivity index data for the Apulia region (Italy): An open-source geodatabase and the innovative CONNECTOSED WebGIS platform.

An open-source geodatabase and its associate WebGIS platform (CONNECTOSED) were developed to collect and utilize data for the Sediment Flow Connectivity Index (SfCI) for the Apulia region of southern Italy. Maps depicting sediment mobility and connectivity across the hydrographic basins of the Apulia region were generated and stored in the geodatabase. This geodatabase is organized into folders containing data in TIFF, shapefile, Jpeg and Pdf formats, including input variables (digital elevation model, land cover map, rainfall map, and soil units dataset for each hydrographic basin), classification graphs (ranking of variable values), dimensionless index maps (slope, ruggedness, rainfall, land cover, and soil stability) and key products (maps of sediment mobility, SfCI, and applied SfCI). The geodatabase maintains the mapping methodology underlying the SfCI algorithm by integrating various Earth datasets from multiple sources through ArcMap™, QGIS® and Matlab® software. This approach aligns surface characteristics with driving forces to describe the spatial variability of sediment pathways and identify hotspot areas. The availability of both input and processed data enables the computation and continuous updating of this applied geomorphological indicator, which is useful for assessing susceptibility to rapid Earth surface changes related to multi-hazard exposure. The geodatabase and the CONNECTOSED platform are valuable tools for researchers and stakeholders involved in land monitoring. The geodatabase and the CONNECTOSED platform are essential tools for researchers, policymakers, and stakeholders involved in land monitoring and environmental management. These tools provide open access to extensive datasets and detailed descriptions of surface dynamics, establishing connections between the causes and effects of extreme phenomena, such as floods, landslides, fires, soil pollution. This integration allows users to combine various forms of environmental data, a capability that is vital for enhancing scientific knowledge, supporting the development of insights, and fostering more informed, evidence-based decision-making in land use planning, conservation efforts, and sustainability initiatives.

Numerical investigation of the velocity field around the spur dike in 90° bend

This study simulated the velocity field in a 90° bend with the spur dike using Computational Fluid Dynamics (CFD) and Artificial Neural Network (ANN). The Model results were compared with the physical model data. The Computational Fluid Dynamics model used governing equations for incompressible viscous fluid motion in the turbulent state. The ANN model used Cartesian coordinate system components as input variables and velocity values as output variables. The results showed that both models could successfully identify the eddy currents around the spur dike, the velocity variations through the channel, the accuracy of the estimating model was assessed by computing the correlation coefficient. The results of the error evaluation criteria, including correlation coefficient (R2), Mean Absolute Percentage Error (MAPE) and Scatter Index (SI) are equal to 0.79, 30.7 and 0.05 for the CFD model and 0.97, 6.1 and 0.02 for the ANN model respectively.

Prediction of dry matter intake in growing Black Bengal goats using artificial neural networks.

Dry matter intake (DMI) determination is essential for effective management of meat goats, especially in optimizing feed utilization and production efficiency. Unfortunately, farmers often face challenges in accurately predicting DMI which leads to wastage of feed and an increase in the cost of production. This investigation aimedto predict DMI in Black Bengal goats by using body weight (BW), body condition score (BCS), average daily gain (ADG), and metabolic body weight (MBW) by applying an artificial neural network (ANN) model. A total of 144 observations were collected from 18 goats over a 4-month period for each input (BW, ADG, MBW and BCS) and output (DMI) variable. These input variables were taken fortnightly and correlated with DMI. The presence of a significant positive correlation between DMI with BW (r = 0.968, p < 0.01), BCS (r = 0.687, p < 0.01), ADG (r = 0.608, p < 0.01), and MBW (r = 0.971, p < 0.01) indicated potential for ANN model development. ANN model with 10 hidden layer neurons trained using GDX and LOGSIG transfer function emergeds as the high-performing model for predicting DMI in Black Bengal goats, achieving the highest R2 (0.9693) and the lowest MSE (0.0013) among the configurations considered. Comparison between the three models revealed that the DMI was estimated more accurately by the ANN model than by linear and second-order non-linear models. ANN may therefore be used to predict DMI with high accuracy and reliability in place of otherregression methods.

Application of Symbolic Classifiers and Multi-Ensemble Threshold Techniques for Android Malware Detection

Android malware detection using artificial intelligence today is a mandatory tool to prevent cyber attacks. To address this problem in this paper the proposed methodology consists of the application of genetic programming symbolic classifier (GPSC) to obtain symbolic expressions (SEs) that can detect if the android is malware or not. To find the optimal combination of GPSC hyperparameter values the random hyperparameter values search method (RHVS) method and the GPSC were trained using 5-fold cross-validation (5FCV). It should be noted that the initial dataset is highly imbalanced (publicly available dataset). This problem was addressed by applying various preprocessing and oversampling techniques thus creating a huge number of balanced dataset variations and on each dataset variation the GPSC was trained. Since the dataset has many input variables three different approaches were considered: the initial investigation with all input variables, input variables with high feature importance, application of principal component analysis. After the SEs with the highest classification performance were obtained they were used in threshold-based voting ensembles and the threshold values were adjusted to improve classification performance. Multi-TBVE has been developed and using them the robust system for Android malware detection was achieved with the highest accuracy of 0.98 was obtained.

Lithium and cobalt recovery from spent battery cathode by acid-reductive leaching using molasses: optimisation by response surface methodology

ABSTRACT The combination of molasses and sulphuric acid is reported for the first time as a leaching agent for lithium and cobalt recovery from spent lithium-ion batteries. The acid-reductive leaching process was optimised using a design of experiment with four input variables: sulphuric acid concentration, molasses concentration, leaching temperature, and leaching time. A response surface model with a central composite design determined the number of experiments, followed by quadratic regression modelling to analyze the relationship between input parameters and leaching recovery. ANOVA assessed the statistical significance of each variable, regression modelling, and optimal leaching conditions. The analysis showed temperature significantly affected lithium recovery, while both temperature and sulphuric acid concentration influenced cobalt recovery. The optimal conditions achieved 97.76% lithium and 99% cobalt recovery (2.5 M sulphuric acid, 60 g/L molasses, 54°C, 108 min). These findings highlight an eco-friendly and efficient approach to metal recovery.

Improvement of rock surface roughness accuracy by combining object space resolution error and 3D point cloud features

To enhance the accuracy of joint roughness coefficient (JRC) estimation in photogrammetry, this study employed a fixed-camera shooting strategy guided by a Structure-from-Motion-based shooting parameter selection algorithm to reconstruct 3D models of rock samples at 16 different shooting distances. The analysis at profile intervals of 0.25 mm, 0.5 mm, and 1 mm revealed a strong correlation between JRC accuracy and three parameters: object space resolution error, spatial distance between point cloud points, and spatial errors of checkpoints on the orientation board. Using these three parameters as input variables and JRC error as the output variable, five machine learning algorithms—Support Vector Regression, Gaussian Process Regression, Multilayer Perceptron, XGBoost, and CatBoost—were employed to predict JRC errors across different shooting distances. The Multilayer Perceptron model performed best at profile intervals of 0.25 mm and 0.5 mm, while XGBoost was optimal at the 1 mm interval. Under the predictions of these models, JRC accuracy improved by an average of 84.7% across the three intervals. Finally, the applicability and limitations of the proposed method were further discussed.

Insights into the specific capacitance of CNT-based supercapacitor electrodes using artificial intelligence.

In this study, the specific capacitance characteristics of a carbon nanotube (CNT) supercapacitor was predicted using different machine learning algorithms, such as artificial neural network (ANN), random forest regression (RFR), k-nearest neighbors regression (KNN), and decision tree regression (DTR), based on experimental studies. The results of the simulation verified the accuracy of the ANN algorithm with respect to the data derived from the specific capacitance of the supercapacitor module. It was observed that there was a strong correlation between the experimental results and the predictions made by the ANN algorithm. Comparative analysis showed that the developed ANN algorithm was consistently superior over other algorithms in terms of different metrics, as indicated by the lowest root mean square error (RMSE) value of roughly 26.24 and the highest R 2 value of approximately 0.91. In contrast, the DTR model recorded the least reliable results in the accuracy analysis, as indicated by the highest RMSE value of about 53.46 and the lowest R 2 value of roughly 0.63. To further explore the impact of independent input parameters including pore structure, specific surface area, and I D/I G ratio on a single output parameter (particularly, the specific capacitance) the sensitivity analysis was also conducted using the SHapley Additive exPlanations (SHAP) framework. This investigation sheds light on the relative significance and effects of different input variables on the specific capacitance of supercapacitors based on CNTs. The results indicated that the ANN algorithm accurately predicted the capacitance of the CNT-based supercapacitor, demonstrating the feasibility and significance of neural network algorithms in the design of energy storage devices.

The Secondary Lifting Performance of Crawler Crane Under Delay Coefficient Control Strategy

Crawler cranes are mobile lifting equipment used in the process of hoisting goods. After the initial lifting, the crane may need a secondary lift due to adjustments in the position or height of the load. Addressing the common issue of load slipping during the secondary lift caused by hydraulic motor reversal, this study proposes a control strategy applicable to crawler crane secondary lifting. Initially establishing the dynamic characteristics of the secondary lift system, incorporating a delay coefficient, and matching motor pressure build-up with memory pressure, the strategy considers a variable pump input current control to identify the relationship between motor pressure build-up and brake release. Analyzing the dynamic characteristics of secondary lifting under different conditions, this study resolves the issue of hydraulic motor reversal during the second lift caused by heavy loads. The results of this study on crawler crane secondary lifting indicate that, when using a delay coefficient of 0.70 and releasing the brake, no slip phenomenon occurred during the secondary lift process under different load conditions, categorized as 200 tons, 600 tons, and 1000 tons. This ensures the stability and transition quality of the secondary lift, providing theoretical guidance for the control of the crawler crane secondary lifting.

Enhancing jointed plain concrete pavement performance analysis through remote data collection on oversize/overweight vehicles: evaluating damage costs and reduced service life

ABSTRACT A large gap exists between actual performance of in-situ Jointed Plain Concrete Pavement (JPCP) and mechanistic models for predicting performance under environmental and traffic loads. JPCP response measurement using instrumented sensors and data acquisition systems is one of the most reliable approaches to bridging the gap. Field instrumentation plays a significant role in validating mechanistic models and establishing correlations between input variables related to JPCP distresses. This paper details the steps and methodology of field instrumentation for installing a remote data-acquisition system in conjunction with JPCP reconstruction conducted in Monona County, Iowa. The data acquisition system installed on-site includes a portable weigh-in-motion system and state-of-the-art sensors for evaluating environmental and traffic impacts on the reconstructed JPCP section by measuring pavement responses related to structural distress. A Finite-Element-Method (FEM)-based JPCP model was also developed and validated using in situ data, ensuring that the model accurately reflects in situ conditions of JPCP and can be reliably used for JPCP analysis under Oversize/Overweight (OSOW) vehicle loads. A life-cycle cost analysis was then conducted to quantify the service-life reduction and the associated costs of JPCP damage due to OSOW vehicle traffic, thereby demonstrating economic implications of accommodating OSOW vehicles on JPCP system.

Precession modulates the poleward expansion of atmospheric circulation to the Arctic Ocean

Under sustained global warming, Arctic climate is projected to become more responsive to changes in North Pacific meridional heat transport as a result of teleconnections between low and high latitudes, but the underlying mechanisms remain poorly understood. Here, we reconstruct subarctic humidity changes over the past 400 kyr to investigate the role of low-to-high latitude interactions in regulating Arctic hydroclimate. Our reconstruction is based on precipitation-driven sediment input variations in the Subarctic North Pacific (SANP), which reveal a strong precessional cycle in subarctic humidity under the relatively low eccentricity variations that dominated the past four glacial-interglacial cycles. Combined with climate model simulations, we highlight that precession drives meridional shifts in the northern rim of the North Pacific Subtropical Gyre (NPSG) and modulates the efficiency of heat and water vapor transfer to the SANP and Arctic regions. Our findings suggest that projections of a northward shift of the NPSG in response to future global warming will lead to wetter conditions in the Arctic Ocean and enhanced sea-ice loss.

In Situ Analysis of Li Plating and Stripping Behaviors Under Dynamic Current Conditions for Realistic Application Scenarios.

Lithium metal batteries are considered the holy grail for next-generation high-energy systems. However, lithium anode faces poor reversibility, unsatisfying cyclability and rate capability due to its uncontrollable plating/stripping behavior. While galvanostatic conditions are extensively studied, the behavior under more realistic application scenarios with variable inputs are less explored. Here, an in situ imaging platform using in-plane microdevice configurations is developed to effectively investigate Li plating/stripping behavior under dynamic conditions. This platform offers high detectivity for analyzing the nuclei size, density, distribution, and growth location, rate, and mode. It is for the first time revealed that nuclei density and growth locations remain constant and are solely determined by the initial nucleation overpotentials during dynamic plating. A transition in growth modes from uniform granular growth to tip-induced dendrite growth, and finally to directional growth among the dendrites is also observed. Guided by these findings, a dynamic plating protocol is proposed, which can greatly improve the Li reversibility and cycling stability. This work not only provides a novel approach to visualize the evolution of key nucleation and growth parameters, especially under variable inputs, but also offers valuable guidance for the future industrialization of metal batteries and the rational design of charging facilities.

Effect of machining parameters on the machinability of 15‐5 precipitation hardened stainless steel during dry turning

AbstractThis research presents the studies on the effect of machining variables and the empirical modeling of machinability output responses during dry turning of 15–5 precipitation‐hardened stainless steel (PHSS) using the taguchi meta‐heuristic algorithm. L9 orthogonal array (OA) robust experimental design was selected for conducting the dry turning operations. The input variables included cutting velocity, depth of cut, and feed rate, while the output responses measured were surface roughness (Ra) and cutting force (Fc). The influence of these process variables was determined through analysis of variance (ANOVA). ANOVA results revealed that cutting speed, feed rate and depth of cut were impacting the average surface roughness by 36 %, 29 %, and 31 %, respectively and influencing the cutting force by 2 %, 16 %, and 72 %, respectively, during turning operation. Empirical models for predicting cutting force and surface roughness were developed using the Taguchi meta‐heuristic algorithm. The optimization process resulted in a significant reduction in surface roughness, with Ra decreasing by 17 % and a notable decrease in cutting force, Fc, by 8 %. These numerical improvements indicate that the proposed optimization approach substantially enhances machining performance, validating its effectiveness.