Prediction Of Carbon Content Research Articles

Just-in-time-learning based prediction model of BOF endpoint carbon content and temperature via vMF mixture model and weighted extreme learning machine

Basic oxygen furnace (BOF) steelmaking is a complicated physical chemical process, in which the endpoint carbon content and temperature are two important indicators. In BOF steelmaking, the quality of raw materials varies greatly between different batches, which would lead to the inaccurate predictions for these two indicators. Additionally, there are imbalance problems in production process data of BOF steelmaking. For the time-varying problem, a novel similarity criterion based on von-Mises Fisher mixture model (VMM) is proposed in this paper and applied for sample selection of just-in-time-learning (JITL)-based endpoint carbon content and temperature prediction model. The V-shaped transfer function is utilized to develop weighted extreme learning machine (WELM) as local regression model to address the imbalance problems. The performance of the proposed methods is compared with other methods under JITL framework. The experimental results show that the proposed online model can provide a more accurate prediction.

Total organic carbon content prediction of source rocks with conventional well log data based on regression committee machine

Total organic carbon content prediction of source rocks with conventional well log data based on regression committee machine

Applying artificial intelligent and empirical techniques for prediction of total organic carbon content from wireline data: A case study from Pabdeh Formation in Mansuri oilfield, SW Iran

Total organic carbon (TOC) content is one of the most crucial parameters needed for petroleum source rock evaluation. In this study, artificial neural network (ANN) and empirically derived ΔlogR techniques were used to estimate organic richness of the Paleocene-Oligocene Pabdeh source rock in the Mansuri oilfield (SW Iran) using the wireline data. The ANN model trained by the Levenberg-Marquardt back propagation algorithm was employed for prediction of TOC content from the responses of the sonic, neutron, density and spectral gamma-ray well logs. The linear regression analysis between TOC values measured by Rock-Eval analysis and the corresponding ones predicted by ΔlogR and ANN methods was performed to validate the performance of each method. The ANN model is capable of predicting TOC contents with higher coefficient of determination (R2) value (0.7830) compared to ΔlogR technique (0.5091). Correspondingly, the mean squared error (MSE) value for the constructed neural network (0.1647) is lower than ΔlogR method (0.2410). Overall, it can be concluded that the proposed ANN model provides more accurate qualitative and quantitative predictions of organic richness, and can be reliably applied for high-resolution estimation of TOC contained in the Pabdeh source rock throughout the Mansuri oilfield.

Prediction Method of TOC Content in Mudstone Based on Artificial Neural Network

Total organic carbon (TOC) content prediction is an important index for evaluating geochemical characteristics of source rocks. Although the methods based on multiple regression and shale content index classification regression can predict TOC content of mudstone in different layers, the models are relatively simple and the results are not accurate. In this paper, the prediction method of artificial neural network is used to predict TOC content of Neogene source rocks in Liping area, Qaidam Basin. The results show that: (1) the correlation between TOC prediction curve and parameters of logging curve is significant; (2) the TOC prediction curve has obvious fluctuation, reflecting the strong heterogeneity of the stratigraphic organic matter; (3) the predicted TOC value is closer to the measured value. Using this method to predict TOC content can be widely used for oil and gas exploration and development in future.

Improving the accuracy of soil organic carbon content prediction based on visible and near-infrared spectroscopy and machine learning

Choosing appropriate multivariate calibration and preprocessing transformation techniques is important in the determination of soil organic carbon (SOC) content based on visible and near-infrared (Vis–NIR) spectroscopy. The performance levels of partial least-squares regression (PLSR), support vector machine regression (SVMR), and wavelet neural network (WNN) calibration methods coupled with different preprocessing approaches were compared using three kinds of criteria, including the coefficient of determination (R2), root mean square error (RMSE), and residual prediction deviation (RPD). A total of 328 soil samples collected from the south bank of Hangzhou Bay were used as the dataset for the calibration–validation procedure and SOC content inversion. The effects of spectra preprocessing transformation methods were evaluated for raw spectra, Savitzky–Golay smoothing with the first derivatives of reflectance (FDR) and Savitzky–Golay smoothing with logarithm of reciprocal of the reflectance (log R−1). The results indicate that the SVMR is superior to the PLSR, and WNN models for SOC content prediction. The combination of the SVMR model with FDR provided the best prediction results for the SOC content, with R2p = 0.92, RPDP = 2.82, RMSEP = 0.36%, and a kappa correlation coefficient of interpolation as high as 0.97. The FDR of Vis–NIR spectroscopy combined with the SVMR model is recommended over the PLSR and WNN modeling techniques for the high-accuracy determination of the SOC content.

A review of end-point carbon prediction for BOF steelmaking process

AbstractThe precise prediction of end-point carbon content in liquid steel plays a critical role in increasing productivity as well as energy efficiency that can be achieved in the basic oxygen furnace (BOF) steelmaking process. Due to numerous and diversity of the studies on BOF end-point carbon prediction, it seems necessary to provide a comprehensive literature review on state-of-the-art developments in end-point carbon prediction for BOF steelmaking. This paper presents the characteristics of different end-point carbon prediction models. The end-point carbon prediction for BOF steelmaking has initially relied on the experience and skill of the operators. With the development of information technology and auto-detection methods, BOF end-point carbon prediction mainly has gone through three stages, such as static prediction, dynamic prediction, and intelligent prediction. Future contributions to the development and application of intelligent end-point carbon prediction in BOF steelmaking are still arduous tasks. However, it is envisaged that the intelligent end-point carbon prediction will witness more frequent applications and greatly improve the high-quality, high-efficiency, and stable production for BOF steelmaking in the future.

Rapid elemental composition measurement of commercial pellets using line-scan hyperspectral imaging analysis

The use of biomass pellets as a renewable energy source is increasing, leading to the need for rapid assessment of biofuel pellet quality for production monitoring. The purpose of this work was to use line-scan near-infrared (NIR) hyperspectral image technology coupled with chemometric tools to assess the elemental components of biomass pellets. The parameters influencing model performance were investigated, i.e. wavelength and spectral pretreatment technique. Either full wavelength or partial wavelength selected using interval successive projections algorithm (iSPA) and interval genetic algorithm (iGA) were investigated. Either raw spectra or pretreated spectra were used for model development. The models were developed using partial least squares regression (PLSR). The most effective model for the prediction of carbon (C), hydrogen (H), and nitrogen (N) content was developed using iGA wavelength selection and standard normal variate (SNV) spectral pretreatment and provided the highest accuracy with a coefficient of determination of prediction set (r2p) and standard error of prediction (SEP) of 0.83 and 1.33%; 0.84 and 0.17%; and 0.90 and 0.098%, respectively. The model could be used for quality assurance. The S content model was poor and not recommended. The relationship between pellet chemical parameters and reflectance characteristics could be used for predicting C, H, and N of biomass pellets.

Dynamic committee machine with fuzzy-c-means clustering for total organic carbon content prediction from wireline logs

The total organic carbon (TOC) content is of great significance to reflect the hydrocarbon-generation potential in shale reservoirs. The well logs were always used to predict the TOC content, but some linear regression methods do not match well with complex data. The neural network method can improve prediction accuracy, but it always generates unstable prediction models. A static committee machine can reduce errors and uncertainties by combining multiple learners, but the weight of integrating learners is difficult to determine. Therefore, a dynamic committee machine with fuzzy-c-means clustering (DCMF) was proposed to predict the TOC content. Experts in the DCMF include Elman neural network, extreme learning machine, and generalized regression neural network. The fuzzy-c-means clustering algorithm was used as the gate network to perform subtasks decomposition and weights calculation based on input data. The subtasks were used to train more adaptive TOC content prediction models, and the weights were transferred to the combiner to integrate all experts’ outputs into final results. The DCMF was applied in two wells located in the Jiumenchong formation in the Qiannan depression, China. The TOC prediction results using the DCMF method are more accurate than the linear regression method, three individual intelligent algorithms, and the static committee machine. The DCMF also provides a new method for weight calculation by mining potential information of input data.

Prediction of Carbon Content in Fly Ash in Ultra Supercritical Generation Unit with Long Short-term Memory Neural Network

The accurate measurement of carbon content in fly ash is the basis of improving the thermal conversion efficiency of boiler and reducing the coal consumption of power generation. Based on the operation data of 1000MW Ultra-supercritical Unit in a thermal power plant, the input parameters of the model were selected by correlation analysis. The Long Short-Term Memory neural network (LSTM) was used to establish a prediction model with the boiler operation parameters as the input and the carbon content of the boiler fly ash as the output. According to the actual operation data of the unit, the prediction performance of the model was trained and tested. The simulation results show that the average prediction error of the LSTM model is 2.8136%, and 86% of the total samples are data with error less than 4.5%, which means it has high accuracy and strong generalization ability.

Reducing moisture effects on soil organic carbon content prediction in visible and near-infrared spectra with an external parameter othogonalization algorithm

Field spectroscopy and other efficient hyperspectral techniques have been widely used to measure soil properties, including soil organic carbon (SOC) content. However, reflectance measurements based on field spectroscopy are quite sensitive to uncontrolled variations in surface soil conditions, such as moisture content; hence, such variations lead to drastically reduced prediction accuracy. The goals of this work are to (i) explore the moisture effect on soil spectra with different SOC levels, (ii) evaluate the selection of optimal parameter for external parameter othogonalization (EPO) in reducing moisture effect, and (iii) improve SOC prediction accuracy for semi-arid soils with various moisture levels by combing the EPO with machine learning method. Soil samples were collected from grassland regions of Inner Mongolia in North China. Rewetting laboratory experiments were conducted to make samples moisturized at five levels. Visible and near-infrared spectra (350–2500 nm) of soil samples rewetted were observed using a hand-held SVC HR-1024 spectroradiometer. Our results show that moisture influences the correlation between SOC content and soil reflectance spectra and that moisture has a greater impact on the spectra of samples with low SOC. An EPO algorithm can quantitatively extract information of the affected spectra from the spectra of moist soil samples by an optimal singular value. A SOC model that effectively couples EPO with random forest (RF) outperforms partial least-square regression (PLSR)-based models. The EPO–RF model generates better results with R2 of 0.86 and root-mean squared error (RMSE) of 3.82 g kg−1, whereas a PLSR model gives R2 of 0.79 and RMSE of 4.68 g kg−1.

Robustness of visible near‐infrared and mid‐infrared spectroscopic models to changes in the quantity and quality of crop residues in soil

AbstractThe robustness of soil organic carbon (SOC) and total nitrogen (TN) content prediction accuracy by visible near‐infrared spectroscopy (visNIRS) and mid‐infrared spectroscopy (MIRS) models after a change in the quantity or quality of crop residues requires investigation. Arable soils (0–20 cm) from 20 locations across Germany were collected, and 0, 2, 4, or 8 g C kg soil−1 of wheat straw (C/N ratio, 54) or clover (C/N ratio, 13) were added. Before and after a 56‐d incubation, dried and ground samples were measured for SOC and TN content and by visNIRS and MIRS. The complete dataset (n = 280) was subdivided into calibration and validation datasets to test the robustness of partial least squares regression models to changes in crop residue quantity and quality in soil. Noise‐reducing data pretreatments included region selection, moving averages, resampling every second data point, and the Savitzky‐Golay algorithm. The MIRS estimates for SOC (7.4–33 g kg−1) had lower root mean squared error of validation (RMSEV = 0.9–2.9 g kg−1) compared with visNIRS (RMSEV = 1.6–7.1 g kg−1). Total N estimates (0.7–2.8 g kg−1) were more comparable for MIRS (RMSEV = 0.1–0.3 g kg−1) and visNIRS (RMSEV = 0.1–1.0 g kg−1). Loadings of partial least squares regression components suggested the predictive mechanisms for SOC and TN were more similar for visNIRS than for MIRS. Differing crop residue quantity or quality in calibration versus validation resulted in biased SOC and TN estimates by visNIRS and MIRS models. However, calibration with a global residue model containing all soils and crop residue quantities and qualities lowered RMSEV for SOC and TN prediction with visNIRS and MIRS, demonstrating the usefulness of this approach.

Improving the Spatial Prediction of Soil Organic Carbon Content in Two Contrasting Climatic Regions by Stacking Machine Learning Models and Rescanning Covariate Space

Understanding the spatial distribution of soil organic carbon (SOC) content over different climatic regions will enhance our knowledge of carbon gains and losses due to climatic change. However, little is known about the SOC content in the contrasting arid and sub-humid regions of Iran, whose complex SOC–landscape relationships pose a challenge to spatial analysis. Machine learning (ML) models with a digital soil mapping framework can solve such complex relationships. Current research focusses on ensemble ML models to increase the accuracy of prediction. The usual ensemble method is boosting or weighted averaging. This study proposes a novel ensemble technique: the stacking of multiple ML models through a meta-learning model. In addition, we tested the ensemble through rescanning the covariate space to maximize the prediction accuracy. We first applied six state-of-the-art ML models (i.e., Cubist, random forests (RF), extreme gradient boosting (XGBoost), classical artificial neural network models (ANN), neural network ensemble based on model averaging (AvNNet), and deep learning neural networks (DNN)) to predict and map the spatial distribution of SOC content at six soil depth intervals for both regions. In addition, the stacking of multiple ML models through a meta-learning model with/without rescanning the covariate space were tested and applied to maximize the prediction accuracy. Out of six ML models, the DNN resulted in the best modeling accuracies, followed by RF, XGBoost, AvNNet, ANN, and Cubist. Importantly, the stacking of models indicated a significant improvement in the prediction of SOC content, especially when combined with rescanning the covariate space. For instance, the RMSE values for SOC content prediction of the upper 0–5 cm of the soil profiles of the arid site and the sub-humid site by the proposed stacking approaches were 17% and 9% respectively, less than that obtained by the DNN models—the best individual model. This indicates that rescanning the original covariate space by a meta-learning model can extract more information and improve the SOC content prediction accuracy. Overall, our results suggest that the stacking of diverse sets of models could be used to more accurately estimate the spatial distribution of SOC content in different climatic regions.

Hyper-spectral inversion of soil organic carbon content under different land use types

Soil spectral information differ across different land use types. Understanding the appropriate modeling methods for different land use types can efficiently and accurately invert soil organic carbon content. We collected 248 samples from forest, cultivated land and orchard in the north-central part of Fengxin County, Jiangxi Province. First, original spectral reflectance curves were reduced noises with Savitzky-Golay (SG) filter. Then 10 nm resampling method was used to reduce data redundancy. We used partial least squares regression (PLSR), support vector machine regression based on grid search method (GRID-SVR) and support vector machine regression based on particle swarm optimization (PSO-SVR) to construct the inversion models of soil organic carbon content. The results showed that when constructing a single land-use type inversion model, RPD of the PLSR method for forest, cultivated land and orchard was 1.536, 1.315 and 1.493 respectively. RPD of GRID-SVR method increased 0.150, 0.183 and 0.502 than that of PLSR method, respectively. The PSO-SVR method had higher accuracy, with RPD being 20.8%, 10.0% and 2.7% higher than GRID-SVR for forest, cultivated land and orchard, respectively. The RPD of forest and orchard were 2.036 and 2.049, which well predicts soil organic carbon. The RPD of cultivated land was 1.647, which can make a rough estimate of soil organic carbon. The PSO-SVR model had the best prediction effect on soil organic carbon of different land use types, with the prediction accuracy of soil organic carbon content in forest and orchard being close and higher than cultivated land. Soil nutrition diffed acorss different land use types, which affect the prediction of soil organic carbon content. Models for inversion of soil organic carbon should be constructed separately for different land use types.

Development of pedotransfer functions by machine learning for prediction of soil electrical conductivity and organic carbon content

The pedotransfer function is a mathematical model used to convert direct soil measurements into known and unknown soil properties. It provides information for modelling and simulation in soil research, hydrology, environmental science and climate change impacts, including investigating the carbon cycle and the exchange of carbon between soils and the atmosphere to support carbon farming. In particular, the pedotransfer function can provide input parameters for landscape design, soil quality assessment and economic optimisation. The objective of the study was to investigate the feasibility of using a generalised pedotransfer function derived with a machine learning method to predict soil electrical conductivity (EC) and soil organic carbon content (OC) for different regional locations in the state of Victoria, Australia. This strategy supports a unified approach to the interpolation and population of a single regional soils database, in contrast to a range of pedotransfer functions derived from local databases with measurement sets that may have limited transferability. The pedotransfer function generation was based on a machine learning algorithm incorporating the Generalized Linear Mixed Model with interactions and nested terms, with Residual Maximum Likelihood estimation, and a predictor-frequency ranking system with step-wise reduction of predictors to evaluate the predictive errors in reduced models. The source of the data was the Victorian Soil Information System (VSIS), which is a database administered for soil information and mapping purposes. The database contains soil measurements and information from locations across Victoria and is a repository of historical data, including monitoring studies. In total, data from 93 projects were available for inputs to modelling and analysis, with 5158 samples used to derive predictors for EC and 1954 samples used to derive predictors for OC. Over 500 models were tested by systematically reducing the number of predictors from the full model. Five-fold cross-validation was used for estimation of model mean-squared prediction error (MSPE) and mean-absolute percentage error (MAPE). The results were statistically significant with only a gradual reduction in error for the top-ranked 50 models. The prediction errors (MSPE and MAPE) of the top ranked model for EC are 0.686 and 0.635, and 0.413 and 0.474 for OC respectively. The four most frequently occurring predictors both for EC and OC prediction across the full set of models were found to be soil depth, pH, particle size distribution and geomorphological mapping unit. The possible advantages and disadvantages of this approach were discussed with respect to other machine learning approaches.

The high-accuracy prediction of carbon content in semi-coke by laser-induced breakdown spectroscopy

LIBS technology is used in the realization process of semi-coke detection.

A modified mathematical model for end-point carbon prediction of BOF based on off-gas analysis

Several models for end-point carbon prediction of BOF(Basic Oxygen Furnace) based on off-gas analysis were studied in this paper. The advantages and disadvantages of the integral model, the exponential decay model and the cubic fitting model were analyzed respectively. Based on analysis of the characteristics of the decarburization rate curve, a new exponential model was established by the introduction of a correction algorithm. The principle of the proposed model involves applying the decarburization rate curve and the descending gradient of the historical heats to obtain the average decarburization curve and reference decarburization efficiency coefficient using the regression fitting method. According to the deviation between the actual and the predicted decarburization curves, the decarburization efficiency coefficient was corrected to improve the prediction accuracy. Plant trials were carried out in a 210 t converter to compare the performance of the mentioned models. The results showed that the new model exhibited better adaptability and higher accuracy than the other ones. The hit ratio of the new model reached more than 90% for the prediction of end-point carbon content within a tolerance of ±0.02%.

Improving prediction of soil organic carbon content in croplands using phenological parameters extracted from NDVI time series data

Mapping the spatial distribution of soil organic carbon (SOC) content or stock is important for climate change studies and land management decisions. When using environmental covariates to map SOC content or stock, variables indicating human activities have drawn growing attentions. Crop species/crop rotations and agricultural management significantly affect the spatial variation of SOC in croplands. For areas where climatic conditions and farming managements are generally consistent in cultivation territory of one crop species, crop phenology largely indicates the crop response to soil. Therefore, phenological parameters incorporating with crop rotation could be effective for mapping soil organic carbon in these areas. In this study, we extracted phenological parameters from Normalized Difference Vegetation Index (NDVI) time series data, and used these variables with crop rotation for predicting topsoil organic carbon content in a cropland area in Anhui province, China. Forty-nine sampling points were collected in field. For these points, there were three crop rotations each with two crop species per year. Twenty-two HJ-1 A/B images for 2010 year with a 30 m resolution were obtained. Eleven phenological parameters for each of the two growing seasons were obtained with a dynamic threshold method. Various combinations of predictive variables were developed based on variable importance and experimented for predicting topsoil organic carbon using random forest. The prediction results were validated using a cross validation approach. Results showed that base levels (given as the average of the left and right minimum values of a time series profile) for both seasons were the most important predictors in this area. Adding both crop rotation and the two phenological parameters to the natural environment variables increased the prediction accuracies by 50% in terms of R2 and 13.4% in terms of root mean square error (RMSE). This study demonstrates the effectiveness of crop phenology in mapping SOC in croplands.

Total organic carbon content prediction based on support-vector-regression machine with particle swarm optimization

Total organic carbon (TOC) is a significant factor to evaluate the hydrocarbon potential of the unconventional reservoir, while it is difficult to obtain the continuous and accurate prediction result of TOC by traditional methods due to the non-linear relationship between TOC content and wireline log data. However, support vector regress machine is an efficient supervised learning artificial intelligence algorithm, and performs well to deal with non-linear relationship dataset. Therefore, in order to explore a more efficient way for TOC content prediction, the support vector regress machine was tried to predict TOC content through wireline log data. To further verify this, some data tests have been conducted on an offshore exploration prior well data in the northern of Beibu Gulf Basin. After analyzing various kernel functions and optimization methods by cross-validation and error analysis, it can be indicated that the Gaussian kernel function and the particle swarm optimization (PSO) algorithm which used the linear decreasing inertia coefficient and mutated particles are suitable for SVR model and model parameter selection, respectively. Finally, the prediction result of SVR with the best model and model parameters is compared with the results of an improved ΔlogR method and multilayer perceptron method. The comparison results show that SVR has a higher correlation coefficient than other methods. In conclusion, the prediction result of SVR can be thought of as an efficient and reliable method for TOC content prediction. Through the continuous prediction of organic carbon content, the qualitative and quantitative evaluation of high-quality source rocks in this area can be realized, which is of great practical significance for determining the exploration field and the search for favorable target reservoirs.

Mathematical modeling for total organic carbon content prediction with logging parameters by neural networks: A case study of shale gas well in South China

The lower Cambrian Niutitang Formation in northwestern Hunan, South China, has already reached its high or over matured stage and is formed with hydrothermal activity and deposition. Thus, it is extremely difficult to predict the total organic carbon (TOC) content accurately by common methods with well-logging data. To solve this problem, we use artificial neural networks for predicting the TOC of the black shales in our study case. We got the input vectors through principal component analysis and based on the relationships and the logging response mechanism between TOC and logging data. In the back-propagation algorithm, some important parameters including the sample databases, the number of hidden layer nodes, transfer function, and weight value adjustment were all optimized correctly in the networks. Then, we built the mathematical model through a large number of learning sample datum and the error function between the actual and expected outputs, and we found that the results are good according to many performance indicators. In testing samples, mean absolute and relative errors are all reduced probably due to the datum ranges and features being focused, but the accuracy also drops when the numbers of participating samples are reduced. Through redefining the [Formula: see text] sample database, we gained more accurate values for the high-TOC source rock. Finally, we think that the results suggest that the method is suitable for shale gas resource exploration under similar geologic conditions and data characteristics.

支持向量机在回归预测有机碳含量中的应用研究——以川东南地区为例

支持向量机在回归预测有机碳含量中的应用研究——以川东南地区为例