Simple Model Research Articles

Modelling the ultimate pullout resistance of geogrids

This paper provides a review of the modelling of geogrid ultimate pullout resistance. Several analytical models were found to disregard the effect of the geogrid stiffness as well as the particle size distribution of the soil. A sensitivity analysis was done on 289 pullout tests to evaluate the significance of these omitted variables as well as other key variables contributing to pullout resistance. It was found that ultimate pullout resistance was the most sensitive to normal stress followed by geogrid length, geogrid stiffness and the friction angle of the soil. By considering geogrid geometry, stiffness and soil grading descriptors, in addition to the standard variables of length, stress and friction angle, regression models of ultimate pullout capacity improved by 15%. In addition, an alternative model to the simplified linear models of FHWA and EBGEO which maintains dimensional consistency while incorporating geogrid stiffness and non-linearity in the model, was proposed. This alternative model performed 32% better than the simplified linear models.

Coupled Multiphysics Modeling of Lithium-Ion Batteries for Automotive Crashworthiness Applications

Abstract Considerable advances have been made in battery safety models, but achieving predictive accuracy across a wide range of conditions continues to be challenging. Interactions between dynamically evolving mechanical, electrical, and thermal state variables make model prediction difficult during mechanical abuse scenarios. In this study, we develop a physics-based modeling approach that allows for choosing between different mechanical and electrochemical models depending on the required level of analysis. We demonstrate the use of this approach to connect cell-level abuse response to electrode-level and particle-level transport phenomena. A pseudo-two-dimensional model and simplified single-particle models are calibrated to electrical–thermal cycling data and applied to mechanically induced short-circuit scenarios to understand how the choice of electrochemical model affects the model prediction under abuse scenarios. These models are implemented using user-defined subroutines on ls-dyna finite element software and can be coupled with existing automotive crash safety models.

Unsteady nonlinear lifting line model for active gust load alleviation of airplanes

AbstractActive gust load alleviation is an important technology for designing future passenger airplanes to be lighter and thus more environmentally friendly. Unsteady Reynolds-averaged Navier–Stokes (URANS) simulations are typically used to accurately calculate gust loads, but because of their high computational cost, they can only be performed at a few selected operating points. In simpler potential theory models, stall is neglected, resulting in loss of accuracy. In this paper, a low-order unsteady aerodynamics wing model is presented, which is able to represent well compressible flow with stall. Furthermore, the model offers the possibility to modularly incorporate actuators, which allows the design and evaluation of active load alleviation systems. The model is based on a conventional unsteady 2D airfoil model including a dynamic stall model. The dynamic stall model requires viscous steady coefficients, e.g. from 2D steady RANS computations. This 2D airfoil model is coupled with a 3D steady-state lifting line model. The model is applied to the LEISA research airplane and extensively validated with URANS results. It performs well in calculating gust loads with and without simultaneous flap deflections, and provides significantly more accurate results in the case of stall than when stall is neglected.

Multi-layered solid element for seismic analysis of rubber bearings considering stress continuity based on finite particle method

Rubber bearings are crucial components in seismic isolation systems to mitigate the damaging effects of earthquakes on structures. Rubber bearings generally consist of alternate layers of steel and rubber bonded together. Simplified two-point spring models omit this layered construction and have limitations in adaptability and programmability, while refined multiple solid models lead to substantial computational costs. This study presents an effective and efficient approach to simulate rubber bearings by using multi-layered solid element based on the finite particle method (FPM). The traditional “Zigzag” shape function is extended into the 3D situation to reflect the serrated deformation within the layered construction. Based on the simplified assumption of stress continuity, an iterative approach is presented for the deformation gradient to achieve the force continuity between layers in dynamic nonlinear scenarios. The multi-layered solid models for both natural rubber bearings (NRBs) and lead rubber bearings (LRBs) are proposed based on this element. The errors of the hysteresis curves of rubber bearings are below 3.1 % between the numerical results obtained by using the multi-layered solid model and the experimental results. The computational speedup ratio of the proposed multi-layered solid model is verified to be 19 relative to the refined multiple solid model. The proposed method is applied to a base-isolated frame structure to demonstrate its effectiveness in structural seismic analysis.

LUIE: Learnable physical model-guided underwater image enhancement with bi-directional unsupervised domain adaptation

Recently, learning-based underwater enhancement (UIE) methods have made considerable progress, significantly benefiting downstream tasks such as underwater semantic segmentation and underwater depth estimation. Most existing unsupervised UIE methods utilize the atmospheric image formation model to decompose underwater images into background color, transmission map, and scene radiance. However, they rely on simplified physical models for estimating the transmission map, over-simplifying its complex formation, which results in imprecise modeling of underwater scattering effects. Additionally, supervised UIE methods heavily depend on synthetic data or ground truth, leading to limited generalization capabilities due to the substantial domain gap presented in different underwater scenarios. To tackle these challenges, we propose a Learnable physical model-guided unsupervised domain adaptation framework for Underwater Image Enhancement, dubbed LUIE. LUIE learns to predict background light, depth, and scene radiance from an underwater image. We incorporate a learnable network to estimate the transmission map based on the predicted depth map. To minimize the inter-domain gap between synthetic and real underwater images, we introduce a bi-directional domain adaptation method that alternates the background light from each domain. Experimental results demonstrate the effectiveness of our proposed method compared to existing approaches, and high-level experiment results validate that our enhanced underwater results. Experiments in real-world settings on underwater ROVs platform with NVIDIA Jetson AGX Xavier further confirm the effectiveness and efficiency of our work.

A programmable microfluidic platform to monitor calcium dynamics in microglia during inflammation

Neuroinflammation is characterized by the elevation of cytokines and adenosine triphosphate (ATP), which in turn activates microglia. These immunoregulatory molecules typically form gradients in vivo, which significantly influence microglial behaviors such as increasing calcium signaling, migration, phagocytosis, and cytokine secretion. Quantifying microglial calcium signaling in the context of inflammation holds the potential for developing precise therapeutic strategies for neurological diseases. However, the current calcium imaging systems are technically challenging to operate, necessitate large volumes of expensive reagents and cells, and model immunoregulatory molecules as uniform concentrations, failing to accurately replicate the in vivo microenvironment. In this study, we introduce a novel calcium monitoring micro-total analysis system (CAM-μTAS) designed to quantify calcium dynamics in microglia (BV2 cells) within defined cytokine gradients. Leveraging programmable pneumatically actuated lifting gate microvalve arrays and a Quake valve, CAM-μTAS delivers cytokine gradients to microglia, mimicking neuroinflammation. Our device automates sample handling and cell culture, enabling rapid media changes in just 1.5 s, thus streamlining the experimental workflow. By analyzing BV2 calcium transient latency to peak, we demonstrate location-dependent microglial activation patterns based on cytokine and ATP gradients, offering insights contrasting those of non-gradient-based perfusion systems. By harnessing advancements in microsystem technology to quantify calcium dynamics, we can construct simplified human models of neurological disorders, unravel the intricate mechanisms of cell-cell signaling, and conduct robust evaluations of novel therapeutics.

Residual Sketch Learning for a Feature-Importance-Based and Linguistically Interpretable Ensemble Classifier.

Motivated by both the commonly used "from wholly coarse to locally fine" cognitive behavior and the recent finding that simple yet interpretable linear regression model should be a basic component of a classifier, a novel hybrid ensemble classifier called hybrid Takagi-Sugeno-Kang fuzzy classifier (H-TSK-FC) and its residual sketch learning (RSL) method are proposed. H-TSK-FC essentially shares the virtues of both deep and wide interpretable fuzzy classifiers and simultaneously has both feature-importance-based and linguistic-based interpretabilities. RSL method is featured as follows: 1) a global linear regression subclassifier on all original features of all training samples is generated quickly by the sparse representation-based linear regression subclassifier training procedure to identify/understand the importance of each feature and partition the output residuals of the incorrectly classified training samples into several residual sketches; 2) by using both the enhanced soft subspace clustering method (ESSC) for the linguistically interpretable antecedents of fuzzy rules and the least learning machine (LLM) for the consequents of fuzzy rules on residual sketches, several interpretable Takagi-Sugeno-Kang (TSK) fuzzy subclassifiers are stacked in parallel through residual sketches and accordingly generated to achieve local refinements; and 3) the final predictions are made to further enhance H-TSK-FC's generalization capability and decide which interpretable prediction route should be used by taking the minimal-distance-based priority for all the constructed subclassifiers. In contrast to existing deep or wide interpretable TSK fuzzy classifiers, benefiting from the use of feature-importance-based interpretability, H-TSK-FC has been experimentally witnessed to have faster running speed and better linguistic interpretability (i.e., fewer rules and/or TSK fuzzy subclassifiers and smaller model complexities) yet keep at least comparable generalization capability.

Reliability of toxicokinetic modelling for PFAS exposure assessment in contaminated water in northern Italy

IntroductionLong-term contamination of tap water and groundwater by perfluoroalkyl and polyfluoroalkyl substances (PFASs) has been documented in the Veneto region of northern Italy. This study aimed to assess the exposure of individuals residing in the contaminated area and to test several toxicokinetic (TK) models of varying complexities to identify an efficient method for predicting perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS) concentrations in human serum using observed data.The ultimate goal is to provide public health officials with guidance on selecting the appropriate TK model for specific contexts, a reliable and rapid tool to support human bio-monitoring (HBM) studies. MethodsTwo simpler empirical TK models and a more complex multi-compartment physiologically based toxicokinetic (PBTK) model were compared with individual and aggregate data from an HBM study. In addition, the PBPK model was modified by adjusting input parameters and introducing new terms into the equations within the original model code. These modifications aimed to optimize the results compared to the original model, with some versions incorporating adjustments to account for the influence of menstruation in women. All models were evaluated to understand their strengths and weaknesses, providing guidance on the appropriate model to use according to specific scenarios. ResultsThe results obtained from the tested models were quite similar, with significant improvements observed only in the modified models. Simpler models also provided satisfactory results in scenarios involving low PFOS serum concentrations and recent exposure cessation. In many cases, predictions demonstrated high accuracy, particularly at the aggregate level and for women. ConclusionsThese findings suggest that environmental protection agencies and health authorities may benefit from employing the tested models at the aggregate level as an initial step in HBM studies, rather than conducting more invasive and expensive screening campaigns.

Free-Drop Experimental and Simulation Study on the Ultimate Bearing Capacity of Stiffened Plates with Different Stiffnesses under Slamming Loads

Differing from previous studies on free-drop tests, this study focuses on the ultimate bearing capacity and failure mechanism of the ship’s bow under slamming loads. A prototype ship’s bow is selected to design two simplified stiffened plates with different stiffeners, and the lateral slamming loads used are equivalent to flare slamming loads. Free-drop tests of the two simplified models are conducted, and the test setups and procedures are provided. The experimental results of slamming pressures and structural responses are obtained. By comparing with the simulation results obtained by Arbitrary Lagrangian-Eulerian (ALE) fluid–structure coupling, the convergence study, symmetry, and independence verifications are carried out. Finally, the dynamic ultimate bearing capacity of stiffened plates with different stiffnesses under lateral slamming loads is studied. The results show that stiffeners enhance the ability of stiffened plates to resist plastic deformation under slamming loads, and T-section stiffeners can provide greater resistance to plastic deformation than other types.

COMPRESSIVE STRENGTH PREDICTION FOR SELF-COMPACTING CONCRETE USING ARTIFICIAL NEURAL NETWORKS

This study investigates the relationship between concrete constituent materials, slump values, density, strain and compressive strength. The database was set in the laboratory by performing a quality assurance test on the constituent materials to ascertain their suitability for concrete work. For material procurement and selection, the British Standard Department of Environment DOE technique of concrete trial mix designs was used. In this study, twelve concrete mixes were utilized. Two simplified models were developed using the feedback network Artificial Neural Networks architecture (ANNs) with R version 4.0.5 and R studio version 1.2.5033. In the two models, an independent layer comprising six nodes and a dependent layer comprising two nodes were taken. Having carried out the sensitivity analysis, the capacity of the developed equation was evaluated in terms of error metrics MSE and RMSE. The grading envelope for river sand showed that the graph fell perfectly into grading envelope zone 2 while other constituent materials tested were all confirmed suitable for concrete works. The results of compressive strength were obtained from varied water-cement ratios used for the concrete trial mix design which ranges between 0.45 to 0.6. The compressive strength results showed that the results increased consistently for each design mix. It is worthy of note that the strength of the mix fell between grades 15, 20, 25 and 35 respectively for both the 7-day and 28-day strength accordingly. Conversely, the ANN models predicted both 7-day and 28-day strength close to the laboratory value. It is concluded that this research has demonstrated economic viability which would enhance overall construction planning efficiency for future researchers, consultants and contractors in the building industries.

Comparative analysis of federated learning algorithms under extreme non-IID conditions for computer vision

Data heterogeneity presents enormous challenges for Federated Learning (FL) where data is characterized by non-independent and identically distributed (non-IID) patterns across participating clients. Although a prominent issue, there is a lack of comprehensive studies regarding the efficacy of FL algorithms under different configurations of hyperparameters and models. Three state-of-the-art algorithms were chosen to be the focus of this paper FedAvg, FedProx, and SCAFFOLD. Experiments of the paper study varying degrees of data imbalance and number of local training steps to see their effects on model training results. The algorithms and hyperparameters were evaluated using ResNet50, AlexNet, and DenseNet121 on the CIFAR-10 dataset. Experimental results reveal that FedAvg and SCAFFOLD generally outperform FedProx, except for the latter excelling in the scenario of extreme non-IID data distributions in combination with minimal local training. It was also observed that models with complex architectures like ResNet50 are more susceptible to data imbalances, while simpler models such as AlexNet prove to be more robust. This study provides valuable insights to how different configurations of hyperparameters affect FL in non-IID scenarios, contributing to future development of more efficient and robust FL algorithms.

Genome-wide discovery for biomarkers using quantile regression at biobank scale

Genome-wide association studies (GWAS) for biomarkers important for clinical phenotypes can lead to clinically relevant discoveries. Conventional GWAS for quantitative traits are based on simplified regression models modeling the conditional mean of a phenotype as a linear function of genotype. We draw attention here to an alternative, lesser known approach, namely quantile regression that naturally extends linear regression to the analysis of the entire conditional distribution of a phenotype of interest. Quantile regression can be applied efficiently at biobank scale, while having some unique advantages such as (1) identifying variants with heterogeneous effects across quantiles of the phenotype distribution; (2) accommodating a wide range of phenotype distributions including non-normal distributions, with invariance of results to trait transformations; and (3) providing more detailed information about genotype-phenotype associations even for those associations identified by conventional GWAS. We show in simulations that quantile regression is powerful across both homogeneous and various heterogeneous models. Applications to 39 quantitative traits in the UK Biobank demonstrate that quantile regression can be a helpful complement to linear regression in GWAS and can identify variants with larger effects on high-risk subgroups of individuals but with lower or no contribution overall.

Comparative analysis of circadian lighting models: melanopic illuminance vs. circadian stimulus

The influence of light exposure on human circadian rhythms has been widely recognized. This effect is mediated by a phototransduction process projected by the intrinsically photosensitive retinal ganglion cells (ipRGCs). The process also involves signal inputs from visual photoreceptors. However, the relative contributions of each photoreceptor to this process remain unclear; accordingly, two different types of circadian lighting models have been proposed: (i) the melanopic illuminance model based solely on ipRGC activation, including melanopic equivalent daylight D65 illuminance (m-EDI) and equivalent melanopic illuminance (EML), and (ii) the circadian stimulus (CS) model, which considers the participation of both ipRGC and visual photoreceptors. However, the two models can yield conflicting predictions. In this study, we assessed and compared the accuracies of these circadian lighting models by fitting a substantial amount of experimental data extracted from multiple laboratory studies. Upon evaluating the results across all exposure durations, data-fitting accuracy of the intricate CS model did not surpass that of the much simpler melanopic illuminance model. Consequently, the latter appears to be the more suitable model for lighting applications. Moreover, a recurring limitation of prior research was revealed: the lighting spectra were not tailored to effectively reflect the fundamental distinctions between the two types of models. Therefore, drawing clear conclusions regarding the accuracies of the models is challenging. To address this problem, we introduced a method for designing contrast-spectra pairs. This method can provide lighting spectra to highlight the difference in circadian illuminance based on one model while keeping the circadian illuminance of others constant.

Feature attention graph neural network for estimating brain age and identifying important neural connections in mouse models of genetic risk for Alzheimer’s disease

Abstract Alzheimer’s disease (AD), a widely studied neurodegenerative disorder, poses significant research challenges due to its high prevalence and complex etiology. Age, a critical risk factor for AD, is typically assessed by comparing physiological and estimated brain ages. This study utilizes mouse models expressing human alleles of APOE and human nitric oxide synthase 2 (hNOS2), replicating genetic risks for AD alongside a human-like immune response. We developed a multivariate model that incorporates brain structural connectomes, APOE genotypes, demographic traits (age and sex), environmental factors such as diet, and behavioral data to estimate brain age. Our methodology employs a Feature Attention Graph Neural Network (FAGNN) to integrate these diverse datasets. Behavioral data are processed using a 2D convolutional neural network (CNN), demographic traits via a 1D CNN, and brain connectomes through a graph neural network equipped with a quadrant attention module that accentuates critical neural connections. The FAGNN model demonstrated a mean absolute error in age prediction of 31.85 days and a root mean squared error of 41.84 days, significantly outperforming simpler models. Our analysis further focused on the brain age delta, which assesses accelerated or delayed aging by comparing brain age, predicted by FAGNN, to the chronological age. A high-fat diet and the presence of the human NOS2 gene were identified as significant accelerators of brain aging in the old age group. Key neural connections identified by FAGNN, such as those between the cingulum, corpus callosum, striatum, hippocampus, thalamus, hypothalamus, cerebellum, and piriform cortex, were found to be significant in the aging process. Validation using diffusion MRI-based metrics, including fractional anisotropy and return-to-origin probability measures across these connections, revealed significant age-related differences. These findings suggest that white matter degradation in the connections highlighted by FAGNN plays a key role in aging. Our findings suggest that the complex interplay of APOE genotype with sex, immunity, and environmental factors modulates brain aging and enhance our understanding of AD risk in mouse models of aging.

Polar Glycerolipids and Membrane Lipid Rafts

Current understanding of the structure and functioning of biomembranes is impossible without determining the mechanism of formation of membrane lipid rafts. The formation of liquid-ordered and disordered phases (Lo and Ld) and lipid rafts in membranes and their simplified models is discussed. A new consideration of the processes of formation of lipid phases Lo and Ld and lipid rafts is proposed, taking into account the division of each of the glycerophospholipids into several groups. Generally accepted three-component schemes for modeling the membrane structure are critically considered. A four-component scheme is proposed, which is designed to more accurately assume the composition of lipids in the resulting Lo and Ld phases. The role of the polar head groups of phospholipids and, in particular, phosphatidylethanolamine is considered. The structure of membrane rafts and the possible absence of a clear boundary between the Lo and Ld phases are discussed.

Long‐range Ising model for regional‐scale seismic risk analysis

AbstractThis study introduces the long‐range Ising model from statistical mechanics to the performance‐based earthquake engineering (PBEE) framework for regional seismic damage analysis. The application of the PBEE framework at a regional scale involves estimating the damage states () of numerous structures, typically performed using fragility function‐based stochastic simulations. However, these simulations often assume conditional independence or employ simplistic dependency models among the of structures, leading to significant misrepresentation of regional risk. The Ising model addresses this issue by converting the available information on binary (safe or failure) into a joint probability mass function, leveraging the principle of maximum entropy. The Ising model offers two main benefits: (1) it requires only the first‐ and second‐order cross‐moments, enabling seamless integration with the existing PBEE framework, and (2) it provides meaningful physical interpretations of the model parameters, facilitating the uncovering of insights not apparent from data. To demonstrate the proposed method, we applied the Ising model to 156 buildings in Antakya, Turkey, using post‐hazard damage evaluation data, and to 182 buildings in Pacific Heights, San Francisco, using simulated data from the Regional Resilience Determination tool. In both instances, the Ising model accurately reproduces the provided information and generates meaningful insights into regional damage. The study also investigates the change in Ising model parameters under varying earthquake magnitudes, along with the mean‐field approximation, further facilitating the applicability of the proposed approach.

Neural network extrapolation to distant regions of the protein fitness landscape

Machine learning (ML) has transformed protein engineering by constructing models of the underlying sequence-function landscape to accelerate the discovery of new biomolecules. ML-guided protein design requires models, trained on local sequence-function information, to accurately predict distant fitness peaks. In this work, we evaluate neural networks’ capacity to extrapolate beyond their training data. We perform model-guided design using a panel of neural network architectures trained on protein G (GB1)-Immunoglobulin G (IgG) binding data and experimentally test thousands of GB1 designs to systematically evaluate the models’ extrapolation. We find each model architecture infers markedly different landscapes from the same data, which give rise to unique design preferences. We find simpler models excel in local extrapolation to design high fitness proteins, while more sophisticated convolutional models can venture deep into sequence space to design proteins that fold but are no longer functional. We also find that implementing a simple ensemble of convolutional neural networks enables robust design of high-performing variants in the local landscape. Our findings highlight how each architecture’s inductive biases prime them to learn different aspects of the protein fitness landscape and how a simple ensembling approach makes protein engineering more robust.

Do jumps matter in discrete-time portfolio optimization?

This paper studies a discrete-time portfolio optimization problem, wherein the underlying risky asset follows a Lévy GARCH model. Besides a Gaussian noise, the framework allows for various jump increments, including infinite-activity jumps. Using a dynamic programming approach and exploiting the affine nature of the model, we derive a single equation satisfied by the optimal strategy, and we show numerically that this equation leads to a unique solution in all special cases. In our numerical study, we focus on the impact of jumps and evaluate the difference to investors employing a Gaussian HN-GARCH model without jumps or a homoscedastic variant. We find that both jump-free models yield insignificant values for the wealth-equivalent loss when re-calibrated to simulated returns from the jump models. The low wealth-equivalent loss values remain consistent for modified parameters in the jump models, indicating extreme market situations. We therefore conclude, in support of practitioners’ preferences, that simpler models can successfully mimic the strategy and performance of discrete-time conditional heteroscedastic jump models.

Analysis of application range of simplified models for field to thermo-field to thermionic emission processes from the cathode

Electron emission plays a dominant role in plasma–cathode interactions and is a key factor in many plasma phenomena and industrial applications. It is necessary to illustrate the various electron emission mechanisms and the corresponding applicable description models to evaluate their impacts on discharge properties. In this study, detailed expressions of the simplified formulas valid for field emission to thermo-field emission to thermionic emission typically used in the numerical simulation are proposed, and the corresponding application ranges are determined in the framework of the Murphy–Good theory, which is commonly regarded as the general model and to be accurate in the full range of conditions of the validity of the theory. Dimensionless parameterization was used to evaluate the emission current density of the Murphy–Good formula, and a deviation factor was defined to obtain the application ranges for different work functions (2.5‒5 eV), cathode temperatures (300‒6000 K), and emitted electric fields (105 to 1010 V·m−1). The deviation factor was shown to be a nonmonotonic function of the three parameters. A comparative study of particle number densities in atmospheric gas discharge with a tungsten cathode was performed based on the one-dimensional implicit particle-in-cell (PIC) with the Monte Carlo collision (MCC) method according to the aforementioned application ranges. It was found that small differences in emission current density can lead to variations in the distributions of particle number density due to changes in the collisional environment. This study provides a theoretical basis for selecting emission models for subsequent numerical simulations.

Comparison of KF-Based Vehicle Sideslip Estimation Logics with Increasing Complexity for a Passenger Car

Nowadays, control is pervasive in vehicles, and a full and accurate knowledge of vehicle states is crucial to guarantee safety levels and support the development of Advanced Driver-Assistance Systems (ADASs). In this scenario, real-time monitoring of the vehicle sideslip angle becomes fundamental, and various virtual sensing techniques based on both vehicle dynamics models and data-driven methods are widely presented in the literature. Given the need for on-board embedded device solutions in autonomous vehicles, it is mandatory to find the correct balance between estimation accuracy and the computational burden required. This work mainly presents different physical KF-based methodologies and proposes both mathematical and graphical analysis to explore the effectiveness of these solutions, all employing equal tire and vehicle simplified models. For this purpose, results are compared with accurate sensor acquisition provided by the on-track campaign on passenger vehicles; moreover, to truthfully represent the possibility of using such virtual sensing techniques in real-world scenarios, the vehicle is also equipped with low-end sensors that provide information to all the employed observers.