Optimal Combination Of Variables Research Articles

Streamflow Simulation with High-Resolution WRF Input Variables Based on the CNN-LSTM Hybrid Model and Gamma Test

Streamflow modelling is one of the most important elements for the management of water resources and flood control in the context of future climate change. With the advancement of numerical weather prediction and modern detection technologies, more and more high-resolution hydro-meteorological data can be obtained, while traditional physical hydrological models cannot make full use of them. In this study, a hybrid deep learning approach is proposed for the simulation of daily streamflow in two mountainous catchments of the Daqing River Basin, northern China. Two-dimensional high-resolution (1 km) output data from a WRF model were used as the model input, a convolutional neural network (CNN) model was used to extract the physical and meteorological characteristics of the catchment at a certain time, and the long short-term memory (LSTM) model was applied to simulate the streamflow using the time-series data extracted by the CNN model. To reduce model input noise and avoid overfitting, the Gamma test method was adopted and the correlations between the input variables were checked to select the optimal combination of input variables. The performance of the CNN-LSTM models was acceptable without using the Gamma test (i.e., with all WRF input variables included), with NSE and RMSE values of 0.9298 and 9.0047 m3/s, respectively, in the Fuping catchment, and 0.8330 and 1.1806 m3/s, respectively, in the Zijingguan catchment. However, it was found that the performance of the model could be significantly improved by the use of the Gamma test. Using the best combination of input variables selected by the Gamma test, the NSE of the Fuping catchment increased to 0.9618 and the RMSE decreased to 6.6366 m3/s, and the NSE of the Zijingguan catchment increased to 0.9515 and the RMSE decreased to 0.6366 m3/s. These results demonstrate the feasibility of the CNN-LSTM approach for flood streamflow simulation using WRF-downscaled high-resolution data. By using this approach to assess the potential impacts of climate change on streamflow with the abundant high-resolution meteorological data generated by different climate scenarios, water managers can develop more effective strategies for managing water resources and reducing the risks associated with droughts and floods.

초등학교 영어과 게임의 효과에 관한 메타분석

Objectives The effects of individual studies on using games in elementary school English classes are synthesized to confirm or deny the individual result against the average effect sizes. Recommendations follow for the design of games in elementary school English classes using an optimal combination of modulated variables. Methods Twenty domestic academic publications in the subject of English education made up the set of papers to be examined. The data were arranged in Excel, and the effect size was determined using the meta-analysis program CMA. According to the composition time and type of the game, organizational form, and the 2022 updated curriculum, the results of comparing and assessing the effects of elementary school English curriculum games were separated into knowledge-understanding, process-function, and value-attitude categories. Results The benefits of learning new words have been established in terms of knowledge and comprehension. In terms of process and function, understanding was more important than expressing, and in terms of value and attitude, interest was more important than attitude and confidence. Regarding the length of the game's configuration, utilizing it for the entire 40 minute class was more efficient than using it only in part (about 10 to 30 minutes). The use of online games, followed by bingo and cards, has the biggest impact in terms of game types. Group sessions, followed by individual and paired tasks, had the most positive impacts on game organization. Conclusions Textbook analyses on attention level improvement(Lee, Jung-Hyeon & Ihm, Hee-Jeong, 2014; Jang, Jin-Kyung & Lee, Seong-Won, 2014; Kim, Sung-Shin & Boo, Kyung-Soon, 2009; Choi, Hye-Wook, 2006), game types(Eun, Ji-Seon & Lee, Je-Young, 2018; Hwang, Hui-Jae, 2006), and pair impact studies(Han, Mi-Yun & Kim, Hye-Ryun, 2014) are only a few examples of the individual or textbook analysis studies that may be seen in earlier research. These individual studies presents limitations to understand the full impacts of using games to teach English in elementary school despite the fact that these researches are significant as individual studies. By synthesizing the results of previous experimental studies, this study comprehensively examined the effects of knowledge, understanding, process function, and value attitude according to the 2022 revised curriculum. An additional significant conclusion was attained as to how the effects of games are varied depending on categories like organization type, duration and types of games in elementary school English classes.

Autonomous Input Voltage Sharing Control and Triple Phase Shift Modulation Method for ISOP-DAB Converter in DC Microgrid: A Multiagent Deep Reinforcement Learning-Based Method

This article proposes a multiagent (MA) deep reinforcement learning (DRL) based autonomous input voltage sharing (IVS) control and triple phase shift modulation method for input-series output-parallel (ISOP) dual active bridge (DAB) converters to solve the three challenges: the uncertainties of the dc microgrid, the power balance problem, and the current stress minimization of the converter. Specifically, the control and modulation problem of the ISOP-DAB converter is formed as a Markov game with several DRL agents. Subsequently, the MA twin-delayed deep deterministic policy gradient (MA-TD3) algorithm is applied to train the DRL agents in an offline manner. After the training process, the multiple agents can provide online control decisions for the ISOP-DAB converter to balance the IVS, and minimize the current stress among different submodules. Without accurate model information, the proposed method can adaptively obtain the optimal modulation variable combinations in a stochastic and uncertain environment. Simulation and experimental results verify the effectiveness of the proposed MA-TD3-based algorithm.

Tree-Species Classification and Individual-Tree-Biomass Model Construction Based on Hyperspectral and LiDAR Data

Accurate classification of tree species is essential for forest resource monitoring, management, and conservation. Based on the classification of tree species, the biomass model at the individual-tree scale of each tree species can be accurately estimated, which can improve the estimation efficiency of individual-tree biomass. In this study, we first extracted four categories of indicators: canopy height model, spectral features, vegetation indices, and texture features from airborne-laser-scanning (ALS) data and hyperspectral data. We used these features as inputs to the random forest algorithm and screened out the optimal variable combination for tree-species classification, with an overall accuracy of 84.4% (kappa coefficient = 0.794). Then, we used ALS data to perform tree segmentation in forest plots to extract tree height, crown size, crown projected area, and crown volume. According to multivariate nonlinear fitting, the parameters of the individual-tree structure were introduced into the constant allometric ratio (CAR) biomass model to establish the biomass models of three tree species: Douglas fir, Red alder, and Bigleaf maple. The results showed that the model-fitting effects were improved after introducing the crown parameters. In addition, we also found that better tree segmentation results led to more accurate structural parameters.

Effect of ultrasound pretreatment and microwave vacuum drying in the production of dried poniol fruit

AbstractThis study suggested a combination of microwave and vacuum drying to meet the requirements of preserving the nutritional content of fresh poniol fruit, consuming lower energy, and having a faster drying rate. In this study, the influence of microwave power, vacuum level, and citric acid pretreatment level on the product quality of dried poniol was analyzed. The citric acid pretreatment was conducted in an ultrasonicator bath. The pretreated was dried by changing the microwave power () from 400 to 800 W and the vacuum level () from 300 to 600 mmHg. The quality of the dried product was analyzed in terms of total phenolic content (), antioxidant activity (), total change in color (), and rehydration ratio () of dried poniol fruit. The response range in the sequence was between 22.219–38.361 mg GAE/g, 69.037–92.578 %DPPH inhibition, 23.451–40.148, and 1.38–3.81, respectively, at various combinations of independent variables. The optimal combination of independent variables for maximal recovery of total phenolic content, antioxidant activity, rehydration ratio, and minimal color changes was obtained at 440 W power level, 600 mmHg vacuum, and pretreated with citric acid of concentration .8%. Under the optimum combination of independent variables, the total phenolic content, antioxidant activity, rehydration ratio, and total color change values were 37.904 mg GAE/g, 92.973%, 23.906, and 3.307, respectively. Because of the low temperature in the vacuum process combined with the rapid heating time of microwaves, the results indicate that microwave vacuum drying might be a viable alternative method for improving the quality of dried food products.Practical ApplicationsMicrowave vacuum drying is a technology that uses microwave radiation to generate heat in the absolute pressure range between the triple point of water and atmospheric pressure. Heat is generated in microwave vacuum drying by directly converting electromagnetic energy into kinetic molecular energy. Hence, heat is generated deep inside the product being dried. Because interior temperatures are greater than exterior temperatures, evaporating water is driven to the surface, resulting in higher partial pressure. As a result, the moisture evaporating within the product is released through the pore structure of the macro‐capillary system of the solid material, resulting in a faster drying rate. Since the superficial layer does not dry out completely, the surfaces remain porous. This method results in a superior end product that is also very nutritious with enhanced sensory properties. As a result, this approach is used in the chemical, pharmaceutical, and food industries.

A Study on Design Variables for Increasing the Breaking Strength of Synthetic Fiber Chain

In this study, a fiber chain was developed to replace a steel chain using high-modulus polyethylene DM20. The pick count, wrapping count, and inner length were selected as the main design variables of the fiber chain and were analyzed to increase the breaking strength. Orthogonal array experiments were conducted, and the results were analyzed with respect to the breaking strength. The analysis revealed that the pick count and wrapping count had meaningful effects at significance levels within 5%. The main effect analysis revealed that a smaller pick count, larger wrapping count, and longer inner length caused the breaking strength to increase. With the wrapping count fixed at 1, a pick count less than -0.65, and an inner length greater than 0.38, the breaking strength was calculated to be greater than 300 kN. These results are expected be important factors in the derivation of an optimal combination of design variables to attain a fiber chain with a targeted strength.

Comparative analysis and optimization of thermodynamic behavior of combined gas-steam power plant using grey-taguchi and artificial neural network

In the published studies, to the best of the authors’ understanding, the grey Taguchi-based statistical technique has not been applied for the optimization of combined gas-steam power plants. In view of this, seven essential input parameters namely compressor inlet air temperature, pressure ratio, fuel temperature, volumetric flow rate of fuel, gas turbine maximum temperature, compressor efficiency, and turbine efficiency are chosen with the aim of determining the optimal combination of design variables that maximize the net power generation, thermal efficiency, exergetic effciency, and minimize the specific fuel consumption. Also, the impact weight of each parameter on output indicators has been evaluated. While the Taguchi approach helps to create an orthogonal array of L27 (3^7), the ANOVA method determines the contribution of each input argument on the objective function. Unlike the Taguchi and ANOVA optimization methodology, the grey relational analysis is performed to transform the multi-objective function into a single objective by way of estimating its grey relational grade. The most favorable combination of input parameters is determined as A1B1C1D1E3F3G3 and under this state, the optimum values of power generation, thermal efficiency, exergetic efficiency, and specific fuel consumption are found to be 259911 kW, 64.9 %, 66.27 %, and 0.1839 kg/kWh respectively. Moreover, the contribution ratio on the output characteristic of the combined cycle is found to be maximum for turbine efficiency (42.41 %) and minimum for fuel temperature (0.59 %). The effectiveness of the grey-Taguchi method is acknowledged and validated using an artificial neural network technique in MATLAB.

Environmental and ecological drivers of harmful algal blooms revealed by automated underwater microscopy

AbstractIn recent years, harmful algal blooms (HABs) have increased in their severity and extent in many parts of the world and pose serious threats to local aquaculture, fisheries, and public health. In many cases, the mechanisms triggering and regulating HAB events remain poorly understood. Using underwater microscopy and Residual Neural Network (ResNet‐18) to taxonomically classify imaged organisms, we developed a daily abundance record of four potentially harmful algae (Akashiwo sanguinea, Chattonella spp., Dinophysis spp., and Lingulodinium polyedra) and major grazer groups (ciliates, copepod nauplii, and copepods) from August 2017 to November 2020 at Scripps Institution of Oceanography pier, a coastal location in the Southern California Bight. Random Forest algorithms were used to identify the optimal combination of environmental and ecological variables that produced the most accurate abundance predictions for each taxon. We developed models with high prediction accuracy for A. sanguinea (), Chattonella spp. (), and L. polyedra (), whereas models for Dinophysis spp. showed lower prediction accuracy (). Offshore nutricline depth and indices describing climate variability, including El Niño Southern Oscillation, Pacific Decadal Oscillation, and North Pacific Gyre Oscillation, that influence regional‐scale ocean circulation patterns and environmental conditions, were key predictor variables for these HAB taxa. These metrics of regional‐scale processes were generally better predictors of HAB taxa abundances at this coastal location than the in situ environmental measurements. Ciliate abundance was an important predictor of Chattonella and Dinophysis spp., but not of A. sanguinea and L. polyedra. Our findings indicate that combining regional and local environmental factors with microzooplankton populations dynamics can improve real‐time HAB abundance forecasts.

Relevance-Based Prediction: A Transparent and Adaptive Alternative to Machine Learning

The authors describe a new prediction system based on relevance, which gives a mathematically precise measure of the importance of an observation to forming a prediction, as well as fit, which measures a specific prediction’s reliability. They show how their relevance-based approach to prediction identifies the optimal combination of observations and predictive variables for any given prediction task, thereby presenting a unified alternative to both kernel regression and lasso regression, which they call CKT regression. They argue that their new prediction system addresses complexities that are beyond the capacity of linear regression analysis but in a way that is more transparent, more flexible, and less arbitrary than widely used machine learning algorithms.

Design optimization of block horn for ultrasonic joining application using RSM–FEA–GA integration approach

This study investigates the design optimization of block horn used in ultrasonic insertion application. Ultrasonic insertion is the process of joining a metal insert with thermoplastic component. As the performance of ultrasonically produced joint depends upon the uniformity of amplitude of vibration generated by horn, the uniformity of displacement amplitude can be improved by optimizing the block horn design. The horn is computationally designed and acoustically analysed by integrating finite element analysis (FEA) with response surface methodology (RSM). The polynomial model for displacement amplitude is developed using RSM. Further, the design of horn is optimized to improve the displacement amplitude by coupling the developed polynomial model with genetic algorithm (GA) as fitness function. The optimized design is acoustically analysed, the results show that the optimal design yields maximum displacement amplitude of 29.57 µm, and the maximum stress induced in the optimal design is 16.3 MPa, which is much lower than fatigue strength of horn material. As the heat developed at joint is quality indicator, the temperature obtained at joint is measured using data acquisition system (DAQ). The measured temperature is 138°C, which correlates well with displacement amplitude developed by optimal design of block horn. The results of this study revealed that the proposed RSM–FEA–GA integration approach managed to find the optimal combination of design variables, which yields maximum displacement amplitude and results good joint performance.

A frequency item mining based energy consumption prediction method for electric bus

For a data-driven bus line energy consumption prediction model, building it with statistical indicators on some variables appeared in the whole bus route, such as average speed, maximum acceleration, etc., always decreases its prediction accuracy due to the discarding of the hidden information in the variable change process. To deal with this problem, a frequency item mining based energy consumption prediction method was proposed, in which the useful prediction information hidden in the process of change is mined by frequency item statistics algorithm and stepwise regression algorithm is used to find the optimal combination of input variables. Simulation and experimental analysis show that with multi-dimensions frequency items, the proposed algorithm can describe and reflect the correlation between different input variables appeared in the process. At the same time, a lot of hardware and software computing costs are saved.

Exploring the Impacts of Data Source, Model Types and Spatial Scales on the Soil Organic Carbon Prediction: A Case Study in the Red Soil Hilly Region of Southern China

Rapid and accurate mapping of soil organic carbon (SOC) is of great significance to understanding the spatial patterns of soil fertility and conducting soil carbon cycle research. Previous studies have dedicated considerable efforts to the spatial prediction of SOC content, but few have systematically quantified the effects of environmental covariates selection, the spatial scales and the model types on SOC prediction accuracy. Here, we spatially predicted SOC content through digital soil mapping (DSM) based on 186 topsoil (0–20 cm) samples in a typical hilly red soil region of southern China. Specifically, we first determined an optimal covariate set from different combinations of multiple environmental variables, including multi-sensor remote sensing images (Sentinel-1 and Sentinel-2), climate variables and DEM derivatives. Furthermore, we evaluated the impacts of spatial resolution (10 m, 30 m, 90 m, 250 m and 1000 m) of covariates and the model types (three linear and three non-linear machine learning techniques) on the SOC prediction. The results of the performance analysis showed that a combination of Sentinel-1/2-derived variables, climate and topographic predictors generated the best predictive performance. Among all variables, remote sensing covariates, especially Sentinel-2-derived predictors, were identified as the most important explanatory variables controlling the variability of SOC content. Moreover, the prediction accuracy declined significantly with the increased spatial scales and achieved the highest using the XGBoost model at 10 m resolution. Notably, non-linear machine learners yielded superior predictive capability in contrast with linear models in predicting SOC. Overall, our findings revealed that the optimal combination of predictor variables, spatial resolution and modeling techniques could considerably improve the prediction accuracy of the SOC content. Particularly, freely accessible Sentinel series satellites showed great potential in high-resolution digital mapping of soil properties.

Remote Sensing Estimation of Forest Aboveground Biomass Based on Lasso-SVR

With the Lutou Forest Farm as the research area, the Lasso algorithm was used for characteristic selection, and the optimal combination of variables was input into the support vector regression (SVR) model. The most suitable SVR model was selected to estimate the aboveground biomass of the forest through the comparison of the kernel function and optimal parameters, and the spatial distribution map of the aboveground biomass in the study area was drawn. The significance analysis of special variables showed good correlations between forest aboveground biomass and each vegetation index. There was a more significant correlation with some remote sensing bands, a less significant correlation with some texture features, and a strong correlation with DEM in the terrain features. When the parameters C is 2 and g is 0.01, the SVR model has the highest precision, which can illustrate 73% of the forest aboveground biomass, with the validation set R2 being 0.62. The statistical analysis of the results shows that the total aboveground biomass of the Lutou Forest Farm is 4.82×105 t. The combination of Lasso with the SVR model can improve the estimation accuracy of forest aboveground biomass, and the model has a strong generalization ability.

Response Surface Methodology Approach to Predict the Flexural Moment of Ferrocement Composites with Weld Mesh and Steel Slag as Partial Replacement for Fine Aggregate

Design of Experiment-Response surface methodology approach is adopted to obtain the optimal flexural moment of ferrocement composites comprising galvanised square weld mesh with weight fraction of fine aggregate by steel slag. To get the optimal combination of progression variables on a flexural moment of ferrocement composites, the central composite design of response surface methodology was adopted. Regression models for responses were justified using analysis of variance and the Pareto chart. The test results show that a maximum ultimate load of 3.30 kN and moment capacity of 220 kNm was obtained for ferrocement with a volume fraction of 2.733% and steel slag of 25% replacement. From the analysis of variance, it is evident that the p value is less than 0.005, the predicted R2 and the adjustable R2 are less than 20%, and the predicted values go in hand with the experimental result which indicates that the proposed models are highly suitable. Moreover, the volume fraction of galvanised square weld mesh has a higher significance on a flexural moment of ferrocement composites. Surface plot, Pareto chart, and regression analysis outcomes show that the most substantial and influential factor for a flexural moment is the volume fraction of galvanised square weld mesh.

Optimization of GuHCl extraction protocol on collagen-based binders in murals by response surface methodology

In this paper, response surface method (RSM) was used to study the effect of extraction parameters on the extracting efficiency of guanidine hydrochloride (GuHCl) on residual protein from mural model samples. The central composite design (CCD) was applied to find the optimal combination of variables, including GuHCl concentration ( x : 1.5-2.5 M), extraction time ( y : 90-180 min), temperature ( z : 50-70 °C) and ultrasound power ( α : 120-280 W) during the extraction. Multiple regression analysis was employed to fit the experimental data into a second-order polynomial equation, and appropriate statistical methods were used for analysis. The optimal extraction conditions were determined as follows: GuHCl concentration of 1.89 M, extraction time of 132 min, temperature of 57 °C, and ultrasonic power of 210 W. The extraction efficiency was 16.685 % under the optimal conditions, which was in agreement with the theoretically predictive value of 17.707 %. The extraction process under optimized parameters had no significant impact on the collagenous structures, characterized by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), ultraviolet (UV) absorption spectroscopy and circular dichroism (CD) spectroscopy.

Prediction of the Nitrogen, Phosphorus and Potassium Contents in Grape Leaves at Different Growth Stages Based on UAV Multispectral Remote Sensing

The rapid and accurate acquisition of nitrogen, phosphorus and potassium nutrient contents in grape leaves is critical for improving grape yields and quality and for industrial development. In this study, crop growth was non-destructively monitored based on unmanned aerial vehicle (UAV) remote sensing technology. Three irrigation levels (W1, W2 and W3) and four fertilization levels (F3, F2, F1 and F0) were set in this study, and drip irrigation fertilization treatments adopted a complete block design. A correlation analysis was conducted using UAV multispectral image data obtained from 2019 to 2021 and the field-measured leaf nitrogen content (LNC), leaf potassium content (LKC) and leaf phosphorus content (LPC) values; from the results, the vegetation indices (VIs) that were sensitive to LNC, LKC and LPC were determined. By combining spectral indices with partial least squares (PLS), random forest (RF), support vector machine (SVM) and extreme learning machine (ELM) machine-learning algorithms, prediction models were established. Finally, the optimal combinations of spectral variables and machine learning models for predicting LNC, LPC and LKC in each grape growth period were determined. The results showed that: (1) there were high demands for nitrogen during the new shoot growth and flowering periods, potassium was the main nutrient absorbed in the fruit expansion period, and phosphorus was the main nutrient absorbed in the veraison and maturity periods; (2) combining multiple spectral variables with the RF, SVM and ELM models could result in improved LNC, LPC and LKC predictions. The optimal prediction model determination coefficient (R2) derived during the new shoot growth period was above 0.65, and that obtained during the other growth periods was above 0.75. The relative root mean square error (RRMSE) of the above models was below 0.20, and the Willmott consistency index (WIA) was above 0.88. In conclusion, UAV multispectral images have good application effects when predicting nutrient contents in grape leaves. This study can provide technical support for accurate vineyard nutrient management using UAV platforms.

Phenological piecewise modelling is more conducive than whole-season modelling to winter wheat yield estimation based on remote sensing data

ABSTRACT Most of the existing remote sensing-based yield estimation methods adopt the mean or cumulative value of meteorological factors within the whole growing season, which may ignore the impact of adverse meteorological conditions on the growth of winter wheat in a certain phenological period. In this study, we distinguished the developmental progression of winter wheat as three phenological periods. In each phenological period, the vegetation indices and meteorological factors were optimized. Then the accuracy and spatiotemporal transferability of the phenological piecewise modelling was compared with that of the whole-season modelling based on four regression methods (i.e. multiple linear regression, artificial neural network, support vector regression and random forest). The results showed that the optimal combinations of variables for the whole-season modelling and the phenological piecewise modelling were different. Compared with the whole-season models, the R2 for the phenological piecewise models improved by 1.4% to 7.6%, the root mean square error (RMSE) decreased by 1.1% to 8.2% among four regression methods . In addition, compared with the whole-season models, the spatiotemporal transferability for the phenological piecewise models was generally better. The accuracies after spatiotemporal transfer for the phenological piecewise models were still higher than that for the whole-season models.

Precipitation variations in the Tai Lake Basin from 1971 to 2018 based on innovative trend analysis

Precipitation is an important variable affecting regional climate characteristics. Accurately identifying trends in precipitation is essential for understanding the evolution of the water cycle in the context of climate change. This study uses an innovative trend analysis (ITA), an innovative polygon trend analysis (IPTA), and wavelet analysis to analyze precipitation at multiple timescales (annual, seasonal, and monthly), and the influencing factors at 30 meteorological stations in the Tai Lake Basin (TLB) from 1971 to 2018. The main conclusions are as follows: 1) the annual precipitation had a significantly increasing trend, while high precipitation had the largest increasing trend, leading to a further increase in the flood risk in the TLB. The precipitation trend mainly decreased in spring and autumn, whereas it mainly increased in summer and winter. Precipitation in different months played crucial but varying roles in the corresponding seasons; there was a sharp transition trend from August to September, whereas the transition from January to February was relatively stable. 2) There was a complex non-linear relationship between precipitation and atmospheric teleconnection. The dominant tele-correlation alone could not explain the relationship between precipitation and large-scale circulation. The highest percentage of significant power includes the optimal combination of variables meant to explain the precipitation variations. 3) The detection results from the ITA method and classic trend analysis methods (Linear regression analysis, Mann Kendall, and Modified Mann Kendall) were consistent; the non-monotonic trends masked by these methods were detectable. IPTA can systematically identify consecutive seasons and monthly transition characteristics as a supplement to ITA. This study can, therefore, provide a reference for water resource management and the prevention and control of droughts and floods.

Machine learning prediction of calorific value of coal based on the hybrid analysis

ABSTRACT As one of the most important indicators of coal, calorific value (CV) not only determines the value of coal product, but also has a significant impact on the further processing and utilization of coal. Traditional methods of obtaining prediction data suffer from a range of problems such as too many input variables, low prediction accuracy of a single analysis method and lack of sensitivity analysis of input variables. In this paper, a novel hybrid analysis was presented to predict the CV of coal. Pearson correlation coefficient (PCC) was used to remove correlations between variables in order to provide a suitable combination of input variables for ML models. The results showed that based on the optimal combination of input variables (ash, Fe, Mg and Na), RF model provided a better regression, better fit and better robustness on the testing set than the other three models when the performance indicators and the number of input variables were considered. In addition, sensitivity analyses of input variables showed the relative importance of individual variables and the way in which each variable affects the output variables. The present work provided novel insights and ideas for understanding the prediction of the CV of coal.

Experimental Investigation of Sorghum Stalk and Sugarcane Bagasse Hybrid Composite for Particleboard

This study evaluated the physical-mechanical properties of three-layer particleboard made from sorghum stalk and sugarcane bagasse hybrid reinforced bonded with urea-formaldehyde resin. The production of particleboards using the sorghum and bagasse particles has been established at the ratio of (25 : 50 : 75) sorghum stalk and sugarcane bagasse with urea-formaldehyde as a matrix at different concentrations (50 kg/m3, 60 kg/m3, and 70 kg/m3) and pressing pressure (18 MPa, 20 MPa, and 22 MPa) through hand mixing from 3 to 5 minutes and then hot press within a temperature range of 160°C to 180°C for 4 minutes. The particleboards were produced with their proportions through the Taguchi design of the experiment (L9) approach. The experimental results were analyzed using Taguchi design and ANOVA with a general linear model. The experimental results showed that internal bonding, flexural strength, modulus of elasticity, and moisture content were significantly improved with high resin content and pressing pressure from 18 MPa to 22 MPa. The thickness swelling value also dropped where the resin concentration varies from 50 kg/m3 to 70 kg/m3 and shows better board stability. In contrast, the water absorption increased as resin content was increased. Results from ANOVA provide that SS3B1P3UFR3 (3 : 1 ratio of Sorghum with bagasse, 22 MPa of pressing load, and 70 kg/m3 urea-formaldehyde concentration) is the most optimal combination of variables for the better performance of particleboard.