Average Absolute Error Research Articles

Algorithms for the global domination problem

A dominating set D in a graph G is a subset of its vertices such that every its vertex that does not belong to set D is adjacent to at least one vertex from set D. A set of vertices of graph G is a global dominating set if it is a dominating set for both, graph G and its complement. The objective is to find a global dominating set with the minimum cardinality. Neither exact nor approximation algorithm existed for the problem known to be NP-hard. We show that it remains NP-hard for restricted types of graphs. At the same time, we specify some families of graphs for which the three heuristics, that we propose here, are optimal. Given the complexity status of the problem, our aim was the development of powerful heuristic algorithms that work well in practice for large-scaled instances. To measure the efficiency of our heuristics, we formulated the problem as an integer linear program (ILP) and also we developed an alternative implicit enumeration (IE) algorithm obtaining guaranteed optimal solutions for the existing benchmark instances with up to 8000 vertices. Remarkably, for 56.75% of these instances, at least one of our heuristics also created an optimal solution, where an average absolute error for the remaining instances was a single vertex. The average approximation ratio was 1.005, whereas for the largest benchmark instances with up to 25000 vertices our heuristics delivered solutions in less than 2 min.

A Hybrid Approach to Modeling Heart Rate Response for Personalized Fitness Recommendations Using Wearable Data

Heart rate (HR) is a key indicator of fitness and cardiovascular health, and accurate HR monitoring and prediction are essential for enhancing personalized fitness experiences. The rise of wearable technology has significantly improved the ability to track personal health, including HR metrics. Accurate modeling of HR response during workouts is crucial for providing effective fitness recommendations, which help users achieve their goals while maintaining safe workout intensities. Although several HR monitoring and prediction models have been developed for personalized fitness recommendations, many remain impractical for real-world applications, and the domain of personalization in fitness applications still lacks sufficient research and innovation. This paper presents a hybrid approach to modeling HR response to workout intensity for personalized fitness recommendations. The proposed approach integrates a physiological model using Dynamic Bayesian Networks (DBNs) to capture heart rate dynamics during workout sessions. DBNs, combined with Long Short-Term Memory (LSTM) networks, model the evolution of HR over time based on workout intensity and individual fitness characteristics. The DBN parameters are dynamically derived from flexible neural networks that account for each user’s personalized health state, enabling the prediction of a full HR profile for each workout, while incorporating factors such as workout history and environmental factors. An adaptive feature selection module further enhances the model’s performance by focusing on relevant data and ensuring responsiveness to new data. We validated the proposed approach on the FitRec dataset, and experimental results show that our model can accurately predict HR responses to workout intensity in future sessions, achieving an average mean absolute error of 5.2 BPM per workout—significantly improving upon existing models. In addition to HR prediction, the model provides real-time fitness personalized recommendations based on individual’s observed workout intensity to an exercise. These findings demonstrate the model’s effectiveness in delivering precise, user personalized heart response to exercise with potential applications in fitness apps for personalized training and health monitoring.

How to rapidly and accurately evaluate the cooling performance of radiative cooling materials?

Radiative cooling materials exhibit great potential for reducing energy consumption in numerous areas. However, owing to the lack of standardized experimental methods and the complexity of model calculation, rapidly and accurately evaluating the cooling performance of radiative cooling materials remains challenging. To address this issue, we propose a new evaluation method that relies solely on the material's radiative properties. First, the calculation accuracies of the hemispherical and spectral-directional emissivity models were compared using the results of three-day field measurements for both radiative cooling and white coatings. The hemispherical emissivity model exhibited slightly higher accuracy and considerably simpler calculation steps than the spectral-directional emissivity model, with the daily, daytime, and nighttime average mean absolute errors between the calculated and measured cooling temperatures all within 2.5 °C. Subsequently, using the hemispherical emissivity model, the cooling performances of radiative cooling materials with shortwave reflectivity (ρsr) and hemispherical emissivity (εlr,he) varied in the range of 0.80–1.00 under different meteorological conditions were calculated. Based on the calculation results, a new concept termed the cooling potential index (CPI) was proposed and calculated as CPI = ρsr + εlr,he/18. The index can be used to rapidly evaluate and compare the cooling performances of different radiative cooling materials.

Predicting the subcutaneous temperature in cryolipolysis using deep operator networks

Accurate monitoring of subcutaneous temperature is crucial for the safety and efficacy of cryolipolysis. However, existing measurement and simulation methods often require trade-offs between accuracy, depth, and computational efficiency. This study introduces a novel deep learning architecture, ConvD-DeepONet, specifically designed to predict subcutaneous temperature fields with both high accuracy and efficiency. The model effectively captures spatial information and produces multi-dimensional output, owing to the innovative integration of convolutional layers and the decoder network. An average absolute error (MAE) of 0.0038 ℃ and a root mean square error (RMSE) of 0.0083 ℃ are achieved, resulting in over a 50 % reduction compared to the baseline models. Moreover, each prediction is completed in just 5.9 ms, rendering it 120 times faster than traditional finite element method simulations. These results indicate that ConvD-DeepONet is a promising tool for real-time subcutaneous temperature prediction, with the potential to enhance the safety and efficacy of cryolipolysis.

A deep learning framework combining molecular image and protein structural representations identifies candidate drugs for pain

Artificial intelligence (AI) and deep learning technologies hold promise for identifying effective drugs for human diseases, including pain. Here, we present an interpretable deep-learning-based ligand image- and receptor's three-dimensional (3D)-structure-aware framework to predict compound-protein interactions (LISA-CPI). LISA-CPI integrates an unsupervised deep-learning-based molecular image representation (ImageMol) of ligands and an advanced AlphaFold2-based algorithm (Evoformer). We demonstrated that LISA-CPI achieved ∼20% improvement in the average mean absolute error (MAE) compared to state-of-the-art models on experimental CPIs connecting 104,969 ligands and 33 G-protein-coupled receptors (GPCRs). Using LISA-CPI, we prioritized potential repurposable drugs (e.g., methylergometrine) and identified candidate gut-microbiota-derived metabolites (e.g., citicoline) for potential treatment of pain via specifically targeting human GPCRs. In summary, we presented that the integration of molecular image and protein 3D structural representations using a deep learning framework offers a powerful computational drug discovery tool for treating pain and other complex diseases if broadly applied.

Predictive modeling of CO2 solubility in piperazine aqueous solutions using boosting algorithms for carbon capture goals

Carbon dioxide (CO2) is the main greenhouse gas that drives global warming, climate change, and other environmental issues. CO2 absorption using amine solvents stands out as one of the most well-known industrial technologies of CO2 capture. However, accurate prediction of CO2 absorption in aqueous amine solutions under different operating conditions is crucial for designing an efficient amine scrubbing system in power plants. In this work, CO2 solubility in aqueous piperazine (PZ) solutions was modeled using 517 experimental data points covering a temperature range of 298 to 373 K, PZ concentration of 0.1 to 6.2 mol/L (M), and CO2 partial pressure of 0.03 to 7399 kPa. To this end, four robust machine learning algorithms, including gradient boosting with categorical features support (CatBoost), light gradient boosting machine (LightGBM), extreme gradient boosting (XGBoost), and adaptive boosting decision trees (AdaBoost-DT) were utilized. Among the developed models, the CatBoost model presented the highest accuracy with an overall determination coefficient (R2) of 0.9953 and an average absolute relative error of 2.36%. Sensitivity analysis revealed that CO2 partial pressure had the greatest influence on CO2 absorption in aqueous PZ solutions, followed by PZ concentration and temperature. Moreover, CO2 partial pressure positively influenced CO2 absorption in aqueous PZ solutions, while PZ concentration and temperature exhibited negative effects. Finally, the leverage technique indicated that both the experimental data bank used for modeling and the model’s estimates were statistically acceptable and valid showing only 8 points (∼1.5% of total data) as possible suspected data.

VSLAM based on deep learning in low-textured scenes

Abstract Traditional Visual Simultaneous Localization and Mapping (VSLAM) algorithms often suffer from the low positioning accuracy and robustness of the traditional feature extraction methods. These methods typically require environments with clear features, perform poorly or even fail in low-textured scenes. To solve this problem, we propose an RGB-D VSLAM algorithm that integrates a deep learning feature extraction network. First, to solve the problem in which sufficient feature points cannot be extracted, the proposed algorithm uses the R2D2 feature extraction network to extract keypoints and generate descriptors. We also perform lightweight optimization on the R2D2 network, ensuring accuracy while doubling the inference speed. Subsequently, we employ the L2 distance to measure the distance between descriptors, thereby enhancing the accuracy of keypoint matching. Ultimately, we add a dense mapping thread to obtain a dense 3D point map generated by the RGB-D information and poses of the keyframes. The experimental results on the TUM dataset show that the average absolute trajectory error is reduced by 59% and the relative trajectory error is reduced by 56%. The results demonstrate that our algorithm significantly improves localization accuracy and system robustness in low-textured environments.

Investigating diverse photogrammetric techniques in the hazard assessment of historical sites of the Museum of the Coal Basin Area in Będzin, Poland

Abstract Assessing the condition of historical sites at risk requires an interdisciplinary approach based on combining multiple measurement technologies. Due to the dynamic development of technology, non-invasive remote sensing methods are gaining significant importance. Among these techniques is photogrammetry based on images taken from unmanned aerial vehicles (UAVs) and those taken with a smartphone. In the study, the authors specified the possibilities and limitations of using remote photogrammetric methods to build accurate digital models of the walls of historic buildings with cracks in them. Point clouds, TIN grids, and façade orthophotos were examined. Statistical analysis was used to determine the repeatability of the data. Two parameters were identified that affect the accuracy of the data: the first – the direction of the segment between two points in the façade plane, and the second – the distance of the segment between two points in the plane of the façade. The study showed that the average accuracy of crack width measurements on the data acquired with the DJI Mavic 3 Enterprise RTK is 1 mm. Testing of crack width measurements using a Samsung Galaxy S20 FE smartphone showed an average absolute error of 0.24 mm. Based on the results, it was concluded that the imagesacquired using mobile devices can be used to determine changes in crack widths on walls.

Billet Size Optimization for Hot Forging of AISI 1045 Medium Carbon Steel Using Zener-Hollomon and Cingara-McQueen Model

This study investigates the effects of initial billet size variations on material flow behavior in hot forging processes, aiming to optimize the forging process using validated predictive models. Material and high-temperature compressive tests inform mathematical models, while simulations are conducted via the finite element method (FEM). Results align with the Zener-Hollomon and Cingara-McQueen approaches. The Arrhenius model predicts AISI 1045 steel flow stress with an R2 of 0.968 and an average absolute relative error (AARE) of 7.079%. The Cingara-McQueen equation achieves an R2 of 0.997 and an AARE of 2.960%. Reducing billets size from 260 mm to 230 mm decreases the material usage by up to 11.5%, while maintaining workpiece integrity. Experimental and simulated loads exhibit an AARE of about 2.69%, thereby indicating potential cost and efficiency improvements in hot forging processes.

Study on microwave ablation temperature prediction model based on grayscale ultrasound texture and machine learning.

Temperature prediction is crucial in the clinical ablation treatment of liver cancer, as it can be used to estimate the coagulation zone of microwave ablation. Experiments were conducted on 83 fresh ex vivo porcine liver tissues at two ablation powers of 15 W and 20 W. Ultrasound grayscale images and temperature data from multiple sampling points were collected. The machine learning method of random forests was used to train the selected texture features, obtaining temperature prediction models for sampling points and the entire ultrasound imaging area. The accuracy of the algorithm was assessed by measuring the area of the hyperechoic area in the porcine liver tissue cross-section and ultrasound grayscale images. The model exhibited a high degree of accuracy in temperature prediction and the identification of coagulation zone. Within the test sets for the 15 W and 20 W power groups, the average absolute error for temperature prediction was 1.14°C and 4.73°C, respectively. Notably, the model's accuracy in measuring the area of coagulation was higher than that of traditional ultrasonic grey-scale imaging, with error ratios of 0.402 and 0.182 for the respective power groups. Additionally, the model can filter out texture features with a high correlation to temperature, providing a certain degree of interpretability. The temperature prediction model proposed in this study can be applied to temperature monitoring and coagulation zone range assessment in microwave ablation.

Chemometrics and analytical blank on the at-line monitoring of Zika-VLP production using near-infrared spectroscopy

The Zika disease caused by the Zika virus was declared a Public Health Emergency by the World Health Union (WHO), with microcephaly as the most critical consequence. Aiming to reduce the spread of the virus, biopharmaceutical organizations invest in vaccine research and production, based on multiple platforms. A crescent vaccine production approach is based on virus-like particles (VLP), for not having genetic material in its composition, hypoallergenic and non-mutant character. For bioprocess, it is essential to have means of real-time monitoring, which can be assessed using process analysis techniques such as Near-infrared (NIR) spectroscopy, that can be combined with chemometric methods, like Partial-Least Squares (PLS) and Artificial Neural Networks (ANN) for prediction of biochemical variables. This work proposes a biochemical Zika VLP upstream production at-line monitoring model using NIR spectroscopy comparing sampling conditions (with or without cells), analytical blank (air, ultrapure water), and spectra pre-processing approaches. Seven experiments in a benchtop bioreactor using recombinant baculovirus/Sf9 insect cell platform in serum-free medium were performed to obtain biochemical and spectral data for chemometrics modeling (PLS and ANN), composed by a random data split (80 % calibration, 20 % validation) for cross-validation of the PLS models and 70 % training, 15 % testing, 15 % validation for ANN. The best models generated in the present work presented an average absolute error of 1.59 × 105 cell/mL for density of viable cells, 2.37 % for cell viability, 0.25 g/L for glucose, 0.007 g/L for lactate, 0.138 g/L for glutamine, 0.18 g/L for glutamate, 0,003 g/L for ammonium, and 0.014 g/L for potassium.

Simplified Neural Network-Based Models for Oil Flow Rate Prediction

Available neural network-based models for predicting the oil flow rate (q<sub>o</sub>) in the Niger Delta are not simplified and are developed from limited data sources. The reproducibility of these models is not feasible as the models’ details are not published. This study developed simplified and reproducible three, five, and six-input variables neural-based models for estimating q<sub>o</sub> using 283 datasets from 21 wells across fields in the Niger Delta. The neural-based models were developed using maximum-minimum (max.-min.) normalized and clip-normalized datasets. The performances and the generalizability of the developed models with published datasets were determined using some statistical indices: coefficient of determination (R<sup>2</sup>), mean square error (MSE), root mean square error (RMSE), average relative error (ARE) and average absolute relative error (AARE). The results indicate that the 3-input-based neural models had overall R<sup>2</sup>, MSE, and RMSE values of 0.9689, 9.6185x10<sup>-4 </sup>and 0.0310, respectively, for the max.-min. normalizing method and R<sup>2</sup> of 0.9663, MSE of 5.7986x10<sup>-3</sup> and RMSE of 0.0762 for the clip scaling approach. The 5-input-based models resulted in R<sup>2</sup> of 0.9865, MSE of 5.7790×10<sup>-4</sup> and RMSE of 0.0240 for the max.-min. scaling method and R<sup>2</sup> of 0.9720, MSE of 3.7243x10<sup>-3</sup> and RMSE of 0.0610 for the clip scaling approach. Also, the 6-input-based models had R<sup>2</sup> of 0.9809, MSE of 8.7520x10<sup>-4</sup> and RMSE of 0.0296 for the max.-min. normalizing approach and R<sup>2</sup> of 0.9791, MSE of 3.8859 x 10<sup>-3</sup> and RMSE of 0.0623 for the clip scaling method. Furthermore, the generality performance of the simplified neural-based models resulted in R<sup>2</sup>, RMSE, ARE, and AAPRE of 0.9644, 205.78, 0.0248, and 0.1275, respectively, for the 3-input-based neural model and R<sup>2</sup> of 0.9264, RMSE of 2089.93, ARE of 0.1656 and AARE of 0.2267 for the 6-input-based neural model. The neural-based models predicted q<sub>o</sub> were more comparable to the test datasets than some existing correlations, as the predicted q<sub>o</sub> result was the lowest error indices. Besides, the overall relative importance of the neural-based models’ input variables on q<sub>o</sub> prediction is S>GLR>P<sub>wh</sub>>T/T<sub>sc</sub>>γ<sub>o</sub>>BS&W>γ<sub>g</sub>. The simplified neural-based models performed better than some empirical correlations from the assessment indicators. Therefore, the models should apply as tools for oil flow rate prediction in the Niger Delta fields, as the necessary details to implement the models are made visible.

An Accurate Critical Total Drawdown Prediction Model for Sand Production: Adaptive Neuro-fuzzy Inference System (ANFIS) Technique

AbstractSand production causes many problems in the petroleum industry. The sand production is predicted to control it in the early stages. Therefore, accurate prediction of sand production has been considered substantial in achieving successful sand control. Critical total drawdown (CTD) can indicate the sand production. The main drawback of the previous studies in predicting CTD is their lack of accuracy. Thus, this study aims to develop an accurate CTD estimation prediction model employing a trend analysis and adaptive neuro-fuzzy inference system (ANFIS). The method is chosen because of its higher performance; the model is built based on 23 published datasets from the Adriatic Sea. The developed ANFIS model is evaluated using various methods, namely, trend analyses. Trend analyses are conducted to show the effects of the features on the CTD to present the physical behavior. The model’s performance was also evaluated using statistical error analyses. In addition, the ANFIS and previously published models were assessed. The trend analyses show the correct relationship between all features and the CTD. In addition, the trend analyses for the previous models are discussed. The results show that the proposed ANFIS method outperforms published methods with an R of 0.9984 and an absolute average percentage relative error (AAPRE) of 4.293%.

YOLOv5-POS: research on cabbage pose prediction method based on multi-task perception technology.

Accurate and rapid identification of cabbage posture is crucial for minimizing damage to cabbage heads during mechanical harvesting. However, due to the structural complexity of cabbages, current methods encounter challenges in detecting and segmenting the heads and roots. Therefore, exploring efficient cabbage posture prediction methods is of great significance. This study introduces YOLOv5-POS, an innovative cabbage posture prediction approach. Building on the YOLOv5s backbone, this method enhances detection and segmentation capabilities for cabbage heads and roots by incorporating C-RepGFPN to replace the traditional Neck layer, optimizing feature extraction and upsampling strategies, and refining the C-Seg segmentation head. Additionally, a cabbage root growth prediction model based on Bézier curves is proposed, using the geometric moment method for key point identification and the anti-gravity stem-seeking principle to determine root-head junctions. It performs precision root growth curve fitting and prediction, effectively overcoming the challenge posed by the outer leaves completely enclosing the cabbage root stem. YOLOv5-POS was tested on a multi-variety cabbage dataset, achieving an F1 score of 98.8% for head and root detection, with an instance segmentation accuracy of 93.5%. The posture recognition model demonstrated an average absolute error of 1.38° and an average relative error of 2.32%, while the root growth prediction model reached an accuracy of 98%. Cabbage posture recognition was completed within 28 milliseconds, enabling real-time harvesting. The enhanced model effectively addresses the challenges of cabbage segmentation and posture prediction, providing a highly accurate and efficient solution for automated harvesting, minimizing crop damage, and improving operational efficiency.

Implementasi Data Mining Prediksi Penjualan Produk Semen Menggunakan Metode Linear Regression (Studi Kasus PT. Toyo Mortar Indonesia)

The cement industry in Indonesia plays a vital role in infrastructure development and the construction sector. PT. Toyo Mortar Indonesia, established in 2015 in the Tangerang industrial area, focuses on the production of instant mortar cement. However, the national cement industry faces the challenge of declining sales due to the pandemic, which has affected people's purchasing power and slowed down property and infrastructure projects. To achieve future growth potential, the right strategy is needed in sales planning. This study implements data mining using the linear regression method with the RapidMiner application to predict cement product sales. The data used includes historical sales data and product information from PT. Toyo Mortar Indonesia. In the tests conducted, the data was processed using 12 variables, of which 10 variables influenced the prediction results: TM-168, TM-178, TM-188, TM-189, TM-198, UM-100, UM-101, UM BIRU, raw materials, and other sales. The test results show that the Linear Regression method can predict cement product sales with an average Root Mean Square Error (RMSE) value of 493,125.701 with a standard deviation of 126,330.525. The average Absolute Error is 303,270.965 with a standard deviation of 46,207.586, and the average relative error is 1.89% with a standard deviation of 0.36%, indicating very good prediction accuracy. This can help the company in making decisions related to sales planning using the available sales data.

The state of charge predication of lithium-ion battery energy storage system using contrastive learning

The state of charge (SOC) is a critical state quantity that must be determined in real-time for a battery energy storage system (BESS). It is a prerequisite for the operation of a BESS. However, obtaining the precise value of SOC is challenging due to it being a hidden state quantity. Existing neural network models commonly employ an end-to-end prediction paradigm for SOC estimation, which fails to fully exploit the rich information present in the time-series battery data. Unlike most studies available in the literature, we propose a novel SOC prediction method named CLDMM. This method is the first to apply contrastive learning techniques from the image field to the SOC prediction of lithium batteries. The method utilizes data augmentation, a multi-scale encoder, and multi-layer perceptrons to learn latent representations and mix these with raw data proportionally for downstream predictive tasks. The performance of the proposed method is evaluated using the Panasonic NCR18650PF dataset, and ablation study were conducted. Experimental results show that CLDMM outperforms baseline methods, achieving an average mean absolute error (MAE) of 0.64% and an average maximum error (MAX) of 2.66%.

New and Highly Accurate Static Young's Modulus Model Using Machine Learning Techniques.

Static Young's modulus (E s) is a critical property required in numerous petroleum calculations. Various models to forecast E s have been proposed in the literature. However, existing models, by and large, lack precision and are confined to specific data set ranges. This study proposes an alternative approach for E s determination, utilizing different machine learning methods, such as an adaptive neuro-fuzzy inference system (ANFIS). In these proposed methods, the predictor variables include bulk formation density (RHOB), shear wave velocity (DTs), and compressional wave velocity (DTc). The models were trained on a data set comprising 1853 hydrocarbon reservoir rock samples from globally diverse locations. They were evaluated using trend, group error, and statistical error analyses. To test the efficacy of the proposed models, the optimally performing model was identified and used to detect the rock types along with the previously published models. Results indicated that ANFIS is the optimum model and can predict E s with an average absolute percentage relative error (AAPRE) of 5.1% and a correlation coefficient (R) of 0.9602. The ANFIS method has some benefits over other machine learning approaches insofar as its superiority in reaching a quicker decision about the mapped relationship between the inputs and outputs because it combines artificial neural networks and fuzzy logic in one tool. The ANFIS can perform a highly nonlinear mapping and displays a better learning ability. The proposed ANFIS model demonstrates its ability to capture accurate physical relationships between input rock properties and E s through trend analysis, which shows that increasing the RHOB increases the E s. Contrarily, increasing the DTc and DTs reduces the E s. Furthermore, the ANFIS model can accurately detect the rock types based on its E s determinations. This research demonstrates the importance of accurately predicting E s for the proper identification of rock types. Thus, this study offers potential advancements in geological assessments of hydrocarbon reservoirs and improvements in many petroleum engineering applications.

Validation and Verification of Computational Model for the Numerical Simulation of the Operational Principle of a Submerged Horizontal Plate Device

This study describes the validation and verification of a two-dimensional numerical wave channel with a Submerged Horizontal Plate (SHP). The SHP is a fixed, submerged structure with a dual function: it acts as a breakwater and as a wave energy converter. To carry out the validation and verification procedures for the computational modeling, experimental and numerical results obtained from the literature were used. A relative error of 2.14% for the maximum axial velocity below the SHP was obtained in the validation, while an average absolute error of 0.0043 m/s for the axial velocity variation below the SHP was identified in the verification. Therefore, it can be inferred that the validation and verification stages were carried out successfully. So, the model is suitable for the numerical simulation of the operational principle of the SHP, both on a laboratory scale and on a real scale, allowing the investigation of its fluid dynamic behavior.

Decoupling and predicting natural gas deviation factor using machine learning methods

Accurately predicting the deviation factor (Z-factor) of natural gas is crucial for the estimation of natural gas reserves, evaluation of gas reservoir recovery, and assessment of natural gas transport in pipelines. Traditional machine learning algorithms, such as Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), Artificial Neural Network (ANN) and Bidirectional Long Short-Term Memory Neural Networks (BiLSTM), often lack accuracy and robustness in various situations due to their inability to generalize across different gas components and temperature-pressure conditions. To address this limitation, we propose a novel and efficient machine learning framework for predicting natural gas Z-factor. Our approach first utilizes a signal decomposition algorithm like Variational Mode Decomposition (VMD), Empirical Fourier Decomposition (EFD) and Ensemble Empirical Mode Decomposition (EEMD) to decouple the Z-factor into multiple components. Subsequently, traditional machine learning algorithms is employed to predict each decomposed Z-factor component, where combination of SVM and VMD achieved the best performance. Decoupling the Z-factors firstly and then predicting the decoupled components can significantly improve prediction accuracy of all traditional machine learning algorithms. We thoroughly evaluate the impact of the decoupling method and the number of decomposed components on the model’s performance. Compared to traditional machine learning models without decomposition, our framework achieves an average correlation coefficient exceeding 0.99 and an average mean absolute percentage error below 0.83% on 10 datasets with different natural gas components, high temperatures, and pressures. These results indicate that hybrid model effectively learns the patterns of Z-factor variations and can be applied to the prediction of natural gas Z-factors under various conditions. This study significantly advances methodologies for predicting natural gas properties, offering a unified and robust solution for precise estimations, thereby benefiting the natural gas industry in resource estimation and reservoir management.

Optimizing battery deployment: Aging trajectory prediction enabling homogenous performance grouping

As battery deployments in electric vehicles and energy storage systems grow, ensuring homogeneous performance across units is crucial. We propose a multi-derivative imaging fusion (MDIF) model, employing advanced imaging and machine learning to predict battery aging trajectories from minimal initial data, thus facilitating effective performance grouping before deployment. Utilizing a derivative strategy and Gramian Angular Difference Field for dimensional enhancement, the MDIF model uncovers subtle predictive features from discharge curve data after only ten cycles. The architecture includes a parallel convolutional neural network with lateral connections to enhance feature integration and extraction. Tested on a self-developed dataset, the model achieves an average root-mean-square error of 0.047 Ah and an average mean absolute percentage error of 1.60%, demonstrating high precision and reliability. Its robustness is further validated through transfer learning on two publicly available datasets, adapting with minimal retraining. This approach significantly reduces the testing cycles required, lowering both time and costs associated with battery testing. By enabling precise battery behavior predictions with limited data, the MDIF model optimizes battery utilization and deployment strategies, enhancing system efficiency and sustainability.