Estimation Of Capacity Research Articles

Traditional matric potential thresholds underestimate soil moisture at field capacity across Oklahoma

AbstractField capacity is a dubious soil physical property, but its use continues because of its perceived value for representing a soil's capacity to store water. Appropriate field capacity estimates can be useful for interpreting data from soil moisture sensors, including those in large‐scale monitoring networks, but suitable methods for defining field capacity in this context are unclear. Motivated by the desire to determine optimal field capacity values for the Oklahoma Mesonet, our objectives were (1) to develop and apply an automated time series analysis algorithm to estimate volumetric soil water content at field capacity and corresponding matric potential and (2) to compare the resulting water contents to those calculated from traditional matric potential thresholds (−33 and −10 kPa). Across 118 Oklahoma Mesonet sites and three soil depths (5, 25, and 60 cm), a matric potential threshold of −10 kPa underestimated field capacity water content by 0.010–0.014 cm cm−3 (3–4%) on average, and a threshold of −33 kPa underestimated it for every site and depth by 0.055–0.078 cm cm−3 (16%−22%) on average. Median matric potentials corresponding to field capacity were −7.6 kPa at the 5‐cm depth, −7.2 kPa at the 25‐cm depth, and −7.3 kPa at the 60‐cm depth. The algorithm developed here can be used to estimate field capacity wherever adequate data are available, and for sites where soil water retention properties are known, matric potentials at field capacity can also be estimated. Using a matric potential of −33 kPa as a standard threshold to represent field capacity is not scientifically justified and should be discontinued.

Enhancing Quality Management of PHC Piles through Improved Driving Formulas: A Comprehensive Review and Analysis

In construction projects, conducting dynamic load tests on all piles proves impractical. Selective testing estimates bearing capacity, while the remaining piles rely on penetration depth for management. This approach, however, faces reliability issues due to varying conditions among piles. Technological advancements, such as non-contact hammers and sensors, have enhanced the accuracy of penetration depth measurements during final driving. Nonetheless, relying solely on penetration depth for construction and quality management remains problematic. This study, therefore, focuses on enhancing the use of driving formulas to improve pile quality management, particularly for the widely used pre-stressed high-strength concrete (PHC) piles. To improve pile quality management, existing driving formulas underwent review and refinement. Utilizing 258 dynamic load test data from various sites, the Hiley, Gates, and Danish formulas underwent validation through statistical analysis and graphical comparison. Enhancements to the Gates formula, achieved through curve fitting with actual data and the application of segment-based coefficients, demonstrated increased accuracy in bearing capacity estimation. These improvements offer a more reliable approach to pile quality management in construction projects.

Pharmacogenetic Variants and Plasma Concentrations of Antiseizure Drugs

Precise estimation of a patient's drug metabolism capacity is important for antiseizure dose personalization. To quantify the differences in plasma concentrations for antiseizure drugs associated with variants of genes encoding drug metabolizing enzymes. PubMed, Clinicaltrialsregister.eu, ClinicalTrials.gov, International Clinical Trials Registry Platform, and CENTRAL databases were screened for studies from January 1, 1990, to September 30, 2023, without language restrictions. Two reviewers performed independent study screening and assessed the following inclusion criteria: appropriate genotyping was performed, genotype-based categorization into subgroups was possible, and each subgroup contained at least 3 participants. The Meta-analysis of Observational Studies in Epidemiology (MOOSE) guidelines were followed for data extraction and subsequent quality, validity, and risk-of-bias assessments. The results from the included studies were pooled with random-effect meta-analysis. Plasma concentrations of antiseizure drugs were quantified with the dose-normalized area under the concentration-time curve, the dose-normalized steady state concentration, or the concentrations after a single dose at standardized dose and sampling time. The ratio of the means was calculated by dividing the mean drug plasma concentrations of carriers and noncarriers of the pharmacogenetic variant. Data from 98 studies involving 12 543 adult participants treated with phenytoin, valproate, lamotrigine, or carbamazepine were analyzed. Studies were mainly conducted within East Asian (69 studies) or White or European (15 studies) cohorts. Significant increases of plasma concentrations compared with the reference subgroup were observed for phenytoin, by 46% (95% CI, 33%-61%) in CYP2C9 intermediate metabolizers, 20% (95% CI, 17%-30%) in CYP2C19 intermediate metabolizers, and 39% (95% CI, 24%-56%) in CYP2C19 poor metabolizers; for valproate, by 12% (95% CI, 4%-20%) in CYP2C9 intermediate metabolizers, 12% (95% CI, 2%-24%) in CYP2C19 intermediate metabolizers, and 20% (95% CI, 2%-41%) in CYP2C19 poor metabolizers; and for carbamazepine, by 12% (95% CI, 3%-22%) in CYP3A5 poor metabolizers. This systematic review and meta-analysis found that CYP2C9 and CYP2C19 genotypes encoding low enzymatic capacity were associated with a clinically relevant increase in phenytoin plasma concentrations, several pharmacogenetic variants were associated with statistically significant but only marginally clinically relevant changes in valproate and carbamazepine plasma concentrations, and numerous pharmacogenetic variants were not associated with statistically significant differences in plasma concentrations of antiseizure drugs.

Variation in Zero Plane Displacement and Roughness Length for Momentum Revisited

Zero plane displacement height (d0\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$d_0$$\\end{document}) and momentum roughness length (z0m\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$z_{0m}$$\\end{document}), describe the aerodynamic characteristics of a vegetated surface. Usually, d0\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$d_0$$\\end{document} and z0m\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$z_{0m}$$\\end{document} are assumed to be constant functions of the physical characteristics of the surface. Prior evidence collected from the literature and our examination of flux tower data show that d0\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$d_0$$\\end{document} and z0m\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$z_{0m}$$\\end{document} vary in time at sites with tree and shrub canopies, but not grasslands. The conventional explanations of these variations are based on linear functions of wind velocity and friction velocity, with little theoretical basis. This study explains the variation in aerodynamic parameters by matching four analytical canopy velocity models to a logarithmic above-canopy velocity profile at canopy height. d0\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$d_0$$\\end{document} and z0m\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$z_{0m}$$\\end{document} come out as functions of 2 non-dimensional terms, the canopy momentum absorption capacity (parameter) and a (measurable) Péclet number. To test the theories of variation, we analysed the velocity profiles from Ozflux and Ameriflux sites. None of the theories could recreate d0\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$d_0$$\\end{document} and z0m\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$z_{0m}$$\\end{document} at half-hourly intervals. However, the canopy velocity models were able better to recreate the distribution of the variations in d0\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$d_0$$\\end{document} and z0m\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$z_{0m}$$\\end{document}. Additionally, the estimates of canopy momentum absorption capacity varied consistently with phenological changes in the canopies, whereas, the fitting parameters of the linear regression of using wind speed and friction velocity did not exhibit physically interpretable variations. The canopy velocity models may offer better predictions with an accurate estimation of the canopy height, a horizontally homogeneous and rigid canopy, and incorporation of the roughness sublayer.

Capacity Estimation for Lithium-Ion Batteries Based on Impedance Spectral Dynamics and Deep Gaussian Process

Capacity Estimation for Lithium-Ion Batteries Based on Impedance Spectral Dynamics and Deep Gaussian Process

A stacking machine learning model for predicting pullout capacity of small ground anchors

Small ground anchors are widely used to fix securing tents in disaster relief efforts. Given the urgent nature of rescue operations, it is crucial to obtain prompt and accurate estimations of their pullout capacity. In this study, a stacking machine learning (ML) model is developed for the rapid estimation of pullout capacity offered by small ground anchors used for temporary tents, leveraging cone penetration data. The proposed stacking model incorporates three ML algorithms as the base regression models: K-nearest neighbors (KNN), support vector regression (SVR), and extreme gradient boosting (XGBoost). A dataset comprising 119 in-situ anchor pullout tests, where the cone penetration data were measured, is utilized to train and assess the stacking model performance. Three metrics, i.e., coefficient of determination (R2), mean absolute error (MAE), and root mean square error (RMSE), are employed to evaluate the predictive accuracy of the proposed model and compare its performance against four popular ML models and an empirical formula to highlight the advantages of the proposed stacking approach. The results affirm that the proposed stacking model outperforms other ML models and the empirical approach as achieving higher R2 and lower MAE and RMSE and more predicted data points falling within 20% error line. Thus, the proposed stacking model holds promising potential as a solution for efficiently predicting the pullout capacity of small ground anchors.

Nationwide cost and capacity estimates for sedimentary basin geothermal power and implications for geologic CO2 storage

Introduction: Sedimentary basins are naturally porous and permeable subsurface formations that underlie approximately half of the United States. In addition to being targets for geologic CO2 storage, these resources could supply geothermal power: sedimentary basin geothermal heat can be extracted with water or CO2 and used to generate electricity. The geothermal power potential of these basins and the accompanying implication for geologic CO2 storage are, however, understudied.Methods: Here, we use the Sequestration of CO2 Tool (SCO2TPRO) and the generalizable GEOthermal techno-economic simulator (genGEO) to address this gap by a) estimating the cost and capacity of sedimentary basin geothermal power plants across the United States and b) comparing those results to nationwide CO2 sequestration cost and storage potential estimates.Results and discussion: We find that across the United States, using CO2 as a geothermal heat extraction fluid reduces the cost of sedimentary basin power compared to using water, and some of the lowest cost capacity occurs in locations not typically considered for their geothermal resources (e.g., Louisiana, South Dakota). Additionally, using CO2 effectively doubles the sedimentary basin geothermal resource base, equating to hundreds of gigawatts of new capacity, by enabling electricity generation in geologies that are otherwise (with water) too impermeable, too thin, too cold, or not deep enough. We find there is competition for the best sedimentary basin resources between water- and CO2-based power, but no overlap between the lowest-cost resources for CO2 storage and CO2-based power. In this way, our results suggest that deploying CO2-based power may increase the cost of water based systems (by using the best resources) and the cost of CO2 storage (by storing CO2 in locations that otherwise may not be targeted). As such, our findings demonstrate that determining the best role for sedimentary basins within the energy transition may require balancing tradeoffs between competing priorities.

Laminarin-modulated osmium nanozymes with high substrate-affinity and selective peroxidase-like behavior engineered colorimetric assay for hydroxyl radical scavenging capacity estimation.

Hydroxyl radical (·OH) scavenging capacity (HOSC) estimation is essential for evaluating antioxidants, natural extracts, or drugs against clinical diseases. While nanozymes offer advantages in related applications, they still face limitations in activity and selectivity. In response, this work showcases the fabrication of laminarin-modulated osmium (laminarin-Os) nanoclusters (1.45 ± 0.05nm), functioning as peroxidase-like nanozymes within a colorimetric assay tailored for rational HOSC estimation. This study validates both the characterization and remarkable stability of laminarin-Os. By leveraging the abundant surface negative charges of laminarin-Os and the surface hydroxyls of laminarin, oxidation reactions are facilitated, augmenting laminarin-Os's affinity for 3,3',5,5'-tetramethylbenzidine (TMB) (KM = 0.04mM). This enables the laminarin-Os-based colorimetric assay to respondto·OH more effectively than citrate-, albumin-, or other polysaccharides-based Os. In addition, experimental results also validate the selective peroxidase-like behavior of laminarin-Os under acidic conditions. Antioxidants like ascorbic acid, glutathione, tannic acid, and cysteine inhibit absorbance at 652nm in the colorimetric platform using laminarin-Os's peroxidase-like activity. Comparedwithcommercial kits, this assay demonstrates superior sensitivity (e.g., responds to ascorbic acid 0.01-0.075mM, glutathione 1-15µg/mL, tannic acid 0.5-5µM, and monoammonium glycyrrhizinate cysteine 1.06-10.63µM) and HOSC testing for glutathione, tannic acid, and monoammonium glycyrrhizinate cysteine. Overall, this study introduces a novel Os nanozyme with exceptional TMB affinity and ·OH selectivity, paving the way for HOSC estimation in biomedical research, pharmaceutical analysis, drug quality control, and beyond.

Co-estimation of state-of-charge and capacity for series-connected battery packs based on multi-method fusion and field data

Accurate state-of-charge (SOC) and capacity estimations are of great importance for the performance management, predictive maintenance, and safe operation of lithium-ion battery packs in electric vehicles (EVs). However, it is quite challenging to estimate real-world large-sized EV battery packs due to the unpredictable operating profiles and large measurement disturbances. This article proposes an adaptive onboard SOC and capacity co-estimation framework, which incorporates a multi-timescale hierarchy and integrates multiple individual methods adaptively to practical driving profiles. First, this framework considers the most evident inconsistency between battery cells and periodically screens the weakest cells in a long timescale (week-level). Subsequently, the SOC of the battery pack is accurately estimated in a short timescale (in real-time) based on multi-method fusion. Finally, the capacity of the battery pack is periodically calibrated in a medium timescale (minute-level) based on an adaptive state filter and reliable SOC estimation. Both the laboratory and field data were used for validation, and the results demonstrated the proposed method achieved accurate SOC and capacity estimations of large-sized EV battery packs, with the maximum root mean squared errors of <0.7 % and <3.2 %, respectively, and it was run five times faster than the multi-cell model-based method.

Probabilistic curvature limit states of corroded circular RC bridge columns: Data-driven models and application to lifetime seismic fragility analyses

Reinforcement corrosion has been recognized as an influential factor in the seismic fragility, both demand and capacity models, of aging reinforced concrete (RC) bridges. For capacity models, accurate and applied prediction tools accounting for aging effects are yet to be well established. Current practices usually perform numerical analyses to obtain time-variant capacity models, which are time-consuming particularly when multi-source structural and environmental uncertainties are considered and sometimes even suffer computational non-convergence. To address these issues, this study leverages a rigorously optimized artificial neural network architecture to develop data-driven models for rapid estimates of probabilistic curvature capacity of corroded circular RC bridge columns with flexural failure modes. An extensive database of multi-level curvature limit states (i.e. slight, moderate, severe, and complete) is created through experimentally validated moment–curvature analyses. A new threshold for the moderate limit state is defined based on the strain of core concrete, rather than cover concrete, to account for the potential full erosion of the cover with drastic corrosion. The data-driven probabilistic capacity models are applied to aid the lifetime seismic fragility assessment of a typical highway bridge, where the spectral acceleration at 1.0 s (Sa-10), peak ground velocity (PGV), and Housner intensity (HI) are found consistently, for the first time, as optimal intensity measures for probabilistic demand modeling of RC columns with different extents of aging effects. For the ease of application, the database and code for the data-driven probabilistic capacity models are accessible at https://bit.ly/3uAa8EY .

An open data-based model for generating a synthetic low-voltage grid to estimate hosting capacity

This study develops and applies an open data-based reference electricity grid analysis (REGAL) model designed to create a synthetic representation of a low-voltage (LV) grid for a country-size geographic area. The model enables large-scale grid simulation in which new loads, such as electric vehicle charging, can be added to estimate their impacts on the current LV grid. The modeling is carried out in three steps: (1) generation of a synthetic LV grid; (2) addition of residential loads, including electric vehicle charging; and (3) evaluating if the grid capacity is exceeded. The grid is generated by selecting transformers and cables so that the system can fulfill the current demand while meeting national regulations and standards for distribution grids, all at the lowest total cost. This paper presents the results of calibration and validation against real-world data for the predicted electricity demands and synthetic grid generated by the model. Different calibration values were explored, and the accuracy of the estimations of grid capacities was calibrated using proprietary real-world data from grid operators. For a region with multiple grid cells, an average deviation from real-world data of ±10 % was achieved. For an average area of 1 km2, the error was 44.5 %, which means that the model is not suitable for analysis on this geographic level. However, the level of accuracy is deemed sufficient for initial estimations of hosting capacity for larger geographic areas, such as a region or a country, thereby enabling estimations of hosting capacity in new areas that lack publicly accessible grid capacities.

Adaptable capacity estimation of lithium-ion battery based on short-duration random constant-current charging voltages and convolutional neural networks

The accurate lithium-ion battery capacity estimation is vital for ensuring the safe and reliable operation of battery-powered systems. Existing data-driven methods heavily rely on fixed charging stages for feature extractions, posing significant limitations in real-world applications. This paper proposes an adaptable capacity estimation approach utilising short-duration random charging voltages during the constant-current charging stage and leveraging convolutional neural networks (CNNs). Based on the user-friendly “Vstart−tend” strategy, two health features including charging voltage and its increment are firstly extracted from random charging segments. Secondly, a feature evolution pattern analysis over the battery's lifespan is proposed to divide the charging voltage range for robust model development. An optimal combination of both the sampling interval and data length is determined for the feature extraction. Then, a two-dimensional CNN model is developed to effectively learn ageing-related knowledge from various random charging segments in a specific charging voltage range. The effectiveness of the proposed approach is ultimately verified using two distinct types of batteries across three operational temperatures. The results demonstrate that the proposed approach show much potential as a promising capacity estimation technique utilising a 600 s random charging segment sampled at a 20 s interval.

A Lower Bound Estimate of the Bearing Capacity of Foundations on Inherently Anisotropic Sands: Verification with 100 Case Studies

A Lower Bound Estimate of the Bearing Capacity of Foundations on Inherently Anisotropic Sands: Verification with 100 Case Studies

Sedimentological and petrophysical characterization of the Bokabil Formation in the Surma Basin for CO2 storage capacity estimation

Large-scale geological sequestration of CO2 is one of the most effective strategies to limit global warming to below 2 °C, as recommended by the Intergovernmental Panel on Climate Change (IPCC). Therefore, identifying and characterizing high-quality storage units is crucial. The Surma Basin, with its four-way dip closed structures, high-quality reservoirs, and thick regional cap rocks, is an ideal location for CO2 storage. This study focuses on the Bokabil Formation, the most prominent reservoir unit in the Surma Basin. Detailed petrographic, petrophysical, XRD, and SEM analyses, along with mapping, have been conducted to evaluate the properties of the reservoir and cap rock within this formation. The Upper Bokabil Sandstone in the Surma Basin ranges from 270 to 350 m in thickness and consists of fine- to medium-grained subarkosic sandstones composed of 70–85% quartz and 5–12% feldspar, with good pore connectivity. Petrophysical analysis of data from four gas fields indicates that this unit has a total porosity of 21–27.4% and a low shale volume of 15–27%. Cross plots and outcrop observations suggest that most of the shales are laminated within the reservoir. The regional cap rock, known as the Upper Marine Shale (UMS), ranges in thickness from 40 to 190 m and contains 10–40 nm nano-type pores. A higher proportion of ductile materials with a significant percentage of quartz in the UMS indicates higher capillary entry pressures, enhancing its capacity to hold CO2. Using the CSLF method with a 6% cut-off of the available pore volume, it is estimated that 103 Mt, 110 Mt, 205 Mt, and 164 Mt of CO2 can be effectively stored in the Sylhet, Kailashtila, Habiganj, and Fenchuganj structures, respectively. Due to the shallow depth of the storage unit and the thick cap rock, the southern Surma Basin is the optimal location for CO2 injection.

Estimation of Groundwater Potential and Aquifer Protective Capacity Within Old Ikenga Hotel, Nsukka, Enugu State, Nigeria

A geophysical survey using electrical resistivity method was conducted around Old Ikenga Hotel, Nsukka in Enugu State, Nigeria to investigate the groundwater potential and aquifer protective capacity of the area. The project area lies within latitudes 6°50'4.0''N – 6°57'52.0"N and longitudes 7°21'6.3"E – 7°28'12.0"E, and covers an area of about 89.6 km2. Vertical Electrical Soundings (VES) were carried out with a digital read out resistivity meter (ABEM SAS 1000). The VES points were marked at 25 m and 75 m along a 100 m line. A total of eight soundings were carried out in the area. The VES data collected were interpreted using INTERPEX software and the results presented in terms of resistivity, thickness, depth and lithology. The lithology was inferred by correlating the result to the lithology log of the borehole drilled in the hotel and the geology of the study area. The VES result shows lithologic layers varying from 4 to 5. Aquiferous sand and Aquiferous sandy shale constitute the aquifer units in the area at depth of 30.26 m to 188.20 m, with their thickness ranging from 30 m to 74 m as shown by their isopach map. The aquifer protective capacity was determined by calculating for longitudinal conductance and matching the values to known standards. The calculated longitudinal conductance varies from 0.0409 to 3.1235 mhos. The interpreted VES results reveal poor, moderate, good and very good aquifer protective capacities of the overburden layers.

Estimation of CO2 storage capacities in saline aquifers using material balance

Estimating CO2 storage capacities is crucial in developing carbon capture and storage projects. Material balance equation (MBE) methods, widely employed for oil and gas reserve estimation, offer a direct approach to estimating CO2 storage capacities. However, previous MBE methods rely on an original fluid in-place volume calculated using volumetric methods to estimate CO2 storage capacities, lacking validation for accuracy. It is essential to accurately estimate the original fluid in-place volume, representing the pore volume, as it substantially influences CO2 storage capacity. This study presents a refined MBE method that ensures accurate estimates of CO2 storage capacities by validating the original fluid in-place volumes in saline aquifers. The accuracy of this method was evaluated by comparing it with a commercial reservoir simulator for a synthetic aquifer example and the Sleipner L9 model. In the synthetic aquifer example, the relative error in CO2 storage capacity estimation with the proposed MBE method was only 2.09%, even when short-term (1-year) injection data were utilized. The proposed MBE method demonstrates consistent accuracy in estimating CO2 storage capacities under different aquifer properties, operating conditions, and MBE-related conditions. The proposed MBE method also accurately estimated the CO2 storage capacity in the Sleipner L9 model, achieving a relative error of 3.47%.

A transfer learning-based ensemble learning model for electric vehicles lithium-ion battery capacity estimation using electrochemical impedance spectroscopy

Lithium-ion battery capacity estimation is crucial to ensure the operational reliability and safety of electric vehicles. Electrochemical Impedance Spectroscopy (EIS) can provide rich physical degradation information of the battery, which makes the EIS-based data-driven method a promising solution for accurate battery capacity estimation. However, batteries tend to present diverse degradation patterns due to the diverse operating conditions and manufacturing, resulting in large capacity estimation errors in practical applications. Therefore, this paper proposes a transfer learning (TL) based ensemble learning (EL) method based on multi-layer perception for the battery capacity estimation. First, the influential impedance features are identified from the EIS using Spearman rank coefficient analysis. Second, three TL models are developed based on distinct TL strategies to enable sufficient knowledge transfer considering the different degradation pattern. Further, an EL method with Gaussian kernel-based weighted assignment technique is proposed to integrate the developed TL models and improve the estimation robustness. The performance of the developed model has been successfully validated only using the first 1/3 cycles of the historical data from target battery obtained under different operating conditions and batteries. The results reveal that the proposed method can largely suppress estimation errors, presenting more accurate capacity estimation compared to other methods with an RMSE lower than 1%.

Unsupervised feature extraction for lithium-ion battery electrochemical impedance spectroscopy and capacity estimation using deep learning method

In this paper, we propose an unsupervised feature extraction technique using a variational autoencoder and generative adversarial network (VAEGAN) that automatically extracts meaningful latent variables from the electrochemical impedance spectra (EIS) of lithium-ion (Li-ion) batteries. These variables accurately reflect the degree of Li-ion battery degradation and are closely related to its capacity. By analyzing the latent variables, it was found that the network can learn the effect of Li-ion battery capacity degradation on the EIS. The extracted latent variables are then used as input features for Gaussian Process Regression (GPR) to estimate the capacity of Li-ion batteries under uneven usage and with unknown historical data. The results show that the capacity prediction method proposed in this paper can significantly reduce prediction errors compared to the current state-of-the-art methods. The method not only provides a more reliable capacity estimation but is also robust to highly noisy EIS data, making it more suitable for practical application scenarios.

Estimation of pile bearing capacity using hybrid models based on modified radial base function

Estimation of pile bearing capacity using hybrid models based on modified radial base function

Data‐driven battery capacity estimation using support vector regression and model bagging under fast‐charging conditions

AbstractLithium‐ion batteries offer significant advantages in terms of their high energy and power density and efficiency, but capacity degradation remains a major issue during their usage. Accurately estimating the remaining capacity is crucial for ensuring safe operations, leading to the development of precise capacity estimation models. Data‐driven models have emerged as a promising approach for capacity estimation. However, existing models predominantly focus on constant current charging conditions, limiting their applicability in real‐world scenarios where fast‐charging conditions are commonly employed. The primary objective of this work is to develop a more versatile machine learning model (i.e., support vector regression [SVR]) capable of estimating battery capacity under fast‐charging conditions, with broader applicability across various work conditions. Genetic algorithm and cross‐validation techniques are employed to simultaneously optimize feature extraction hyperparameters and SVR hyperparameters. A model bagging method is further implemented to address prediction challenges under unknown fast‐charging conditions. The effectiveness of the developed model is validated on a cycling dataset of lithium‐ion batteries under different two‐stage fast‐charging conditions.