Regular Model Research Articles

An intelligence expert system for quality health monitoring of power transformers using testing attributes

Power transformers are very vital components in electrical power network. Deterioration of transformer’s insulation has significant impacts on its health condition. Hence, detailed analysis of insulation testing attributes is very important in diagnosing the transformer. It is possible with an effective decision-making system. In the present paper, a new grey relational-based fuzzy logic expert system is proposed for continuous health monitoring of power transformers. In this approach, ranking-based qualitative relational analysis is initially carried out in order to reduce large number of input attributes into three input health grades. These input grades are further integrated with a technically designed expert fuzzy rule base. Outputs of the proposed model are validated with the outputs of multi-criterion based fuzzy logic model presented in [Mharakurwa E T and Goboza R 2019 Multiparameter-based fuzzy logic health index assessment for oil-immersed power transformers Adv. Fuzzy Syst. 2019]. Testing data of 200 in-service transformers are utilised in the validation of both the models. It is observed from the comparison that similarity index of the proposed grey-fuzzy system is 92 percent. Several issues exist in regular transformer health assessment models such as the large number of rules, uncertainties, requirement of large training data, complexity issues are overcome by integrating the grey relational analysis with the fuzzy logic system. The proposed model can be easily implemented by all utilities and diagnostic experts of transformers.

Smoothing fast proximal gradient algorithm for the relaxation of matrix rank regularization problem

This paper proposes a general inertial smoothing proximal gradient algorithm for solving the Capped-ℓ1 exact continuous relaxation regularization model proposed by Yu and Zhang (2022) [29]. The proposed algorithm incorporates different extrapolations into the gradient and proximal steps. It is proved that, under some general parameter constraints, the singular values of any accumulation point of the sequence generated by the proposed algorithm have the common support set, and the zero singular values can be achieved in a finite number of iterations. Furthermore, any accumulation point is a lifted stationary point of the relaxation model. Numerical experiments illustrate the efficiency of the proposed algorithm on synthetic and real data, respectively.

Simplified regular solution model for asphaltenes precipitation in crude oil blends

The regular solution model is used with Hankinson-Brobst-Thomson (HBT) molar volume correlation for saturates, aromatics, resins and asphaltenes (SARA) fractions for pure crude oils and blends to predict the precipitation of asphaltenes. The parameters of the HBT correlation were tuned with experimental data for SARA fractions densities. At the same temperature and pressure conditions, the precipitation of asphaltenes is greatly influenced by the presence of other compounds. SARA fractions were obtained according with ASTM D2007 using n-heptane as precipitant solvent. The use of average properties for saturates, aromatics and resins and the assumptions made in the model could reproduce precipitation curves for all study cases. Calculations of asphaltene precipitation prediction for pure oils showed an average absolute deviation between 0.12 and 0.69% for oil of Gulf of Mexico and Cold Lake oil, respectively. For blends it showed values between 0.08 and 0.65% for Cold Lake/Lloydminster-Asphaltene free blend and Cold Lake/Lloydminster-Gulf of Mexico blend respectively. The proposed model maintains its simplicity and can predict the asphaltene precipitation accurately.

Proof Load Testing Method by the American Association of State Highway and Transportation Officials and Suggestions for Improvement

Because of the aging of infrastructure, methods are explored by which the reliability of existing bridges and viaducts can be assessed. In cases in which limited information of the structure is available or its condition is of concern, proof load testing may be used to demonstrate sufficient live load carrying capacity. Proof load tests in the U.S.A. are typically performed using the Manual for Bridge Evaluation (MBE) published by the American Association of State Highway and Transportation Officials (AASHTO). The proof load is expressed by the regular live load model magnified by the target proof load factor. The level of reliability obtained using the target proof load factor is not explicitly stated in the MBE, but is of particular interest. In this article, relevant background documents are investigated to uncover the underlying calculations, assumptions, and input data. Current challenges in proof load testing are described in which the considerations of time dependence, stop criteria, available information, and system-level assessment are highlighted. Subsequently, improvements to the MBE proof load testing background are suggested. An example calculation using traffic data from the Netherlands shows that the HL93 load model and Eurocode LM1 provide a reasonably constant proof load factor with span length for bending and shear. However, the HS20 load model does not scale well with increasing span length. It is found that the magnitude of the target load as specified through the proof load factor is directly related to the desired level of reliability. Although the MBE proof load testing method is practical, several challenges remain.

Numerical and Experimental Analyses of a Partially Water-Filled Inclined Floating Body

Abstract Floating bodies are widely used in the field of offshore engineering. Existing studies show that the motion responses of a floating body in waves will change with the internal water in the cabins, and it is essential to analyze its hydrodynamic performance under various potential operating conditions. However, most of the research only considers the interaction between the floating body and the internal water in the upright position, and there has been little research on the inclined floating body caused by water partially filled in the broadside. In this study, a floating body with a plurality of longitudinal and transverse cabins was designed. The regular wave model test was carried out in a wave basin, and the numerical results were compared with the experimental results, which verified the accuracy of the model. The effects of wave direction, wave frequency, water-filling depth, and cabin division on the motion responses of the floating body are analyzed. The results show that the water inside the cabins has a significant impact on the roll motion. With the increase of the water-filling depth, the natural frequency of the roll motion decreases. Special attention should be paid to the impact on the wave direction and cabin division on the partially water-filled inclined floating body.

A CauchyTV non-convex regularization model for MRI reconstruction

A CauchyTV non-convex regularization model for MRI reconstruction

MODELING INTERREGIONAL PATIENT MOBILITY: THEORY AND EVIDENCE FROM SPATIALLY EXPLICIT DATA

AbstractThis article provides theory and evidence on the spatial determinants of regional patient flows. We develop a theoretical model that explains a patient's choice to consult a general practitioner by a measure of spatial accessibility. We empirically test this gravity‐type model using regional patient flows and detailed data on the spatial distribution of residents and physicians in Austria. Our measure of spatial accessibility is a crucial determinant of patient flows that substantially increases the explanatory power of regular gravity models. Counterfactual simulations show heterogeneous effects of exiting physicians on health‐care accessibility and patient mobility.

Evaluation of feature selection methods utilizing random forest and logistic regression for lung tissue categorization using HRCT images

AbstractCategorization of lung tissue patterns with interstitial lung diseases (ILD) utilize high‐resolution computed tomography (HRCT) lung images of the TALISMAN dataset which is challenging due to high intra‐class variation and inter‐class ambiguity. To tackle this, major contributions are made in three aspects. First, a novel shape‐based feature is proposed to quantify the amount of fibrotic and nodular components in a lung tissue pattern which helps to minimize intra‐class variation and inter‐class ambiguity. Second, we address the curse of dimensionality which often arises due to huge feature space. Third, to prevent an overfitting issue, the Grid Search optimization algorithm is utilized by tuning the Random Forest hyper‐parameters. In this manuscript, a framework is proposed to categorize lung tissue patterns by integrating four types of feature domains (a) intensity‐based, (b) texture‐based, (c) wavelet‐based, and (d) shape‐based along with the novel shape‐based feature. As a result, we encounter a large feature space (i.e., ), which leads to high dimensionality. To address this issue, we reduce the feature space using filter‐based f‐statistic, reliefF, minimum Redundancy Maximum Relevance (mRMR), and embedded‐based decision trees, regularization models. We found that the regularization model shrinks the feature space by 2.5 times in just 90 s whereas mRMR methods reduce the feature space by 10 times in 13 min. Using the proposed feature set, we employ Random Forest and Logistic Regression as potential classifiers to classify lung tissue patterns. Experiential results reveal that the proposed framework categorizes lung tissue patterns more effectively than state‐of‐the‐art hand‐crafted and deep learning‐based approaches.

An iteratively reweighted time-domain acoustic method for reconstructing the transient acoustic field

An improved iterative reweighted least-squares based on real-time near-field acoustic hologram method (RTNAH) is proposed to reconstruct the transient acoustic field. First, time-domain convolution relation formula between the pressure time-wavenumber spectra on the hologram and reconstruction planes is first established by an impulse response function, and the corresponding matrix equation is obtained. Then, for solving the pressure spectra on the reconstruction plane, an improved iterative reweighted least-squares algorithm is applied for solving the minimization problem of the sum of 2-norm residual and 2-norm weighted solution under the weighted and sparse representation framework. By controlling the iteratively updated weighting matrix, the sparsity of the solution is promoted. Finally, the pressure spectra of the reconstruction plane are estimated, and the corresponding time-varying pressures are acquired by the two-dimensional inverse Fourier transform. A numerical simulation is conducted to observe the reconstruction capacity of the proposed method. The simulation results demonstrate that the proposed method can reconstruct the transient acoustic field in time-space domains very well. Meanwhile, several vital parameters are discussed, and the superiorities of the proposed method are testified by comparing to RTNAH with Tikhonov regularization and YALL1 model. An experiment is implemented for further validating the effectivity of the proposed method.

Regularized SPH model for soil–structure interaction with generalized frictional boundary

AbstractThis paper proposes a regularized smoothed particle hydrodynamics (SPH) scheme to solve the problems of tensile instability and stress noise in large‐deformation soil–structure interaction. Particle shifting and stress regularization are combined to maintain uniform stresses and particle distributions. A generalized frictional boundary condition is presented to model soil–structure interfaces with abitrary friction coefficients. A modified particle shifting technique is developed to regularize particles near the frictional boundaries. Furthermore, GPU acceleration algorithm is adopted to improve the efficiency. The proposed scheme is compared with the conventional SPH methods using numerical examples, including the collapse of three‐dimensional granular column, the collapse of the cohesive soil, and pipe penetrates into soft clay. The comparison indicates that the proposed regularized SPH model can give highly accurate results for large‐deformation soil–structure interaction problems with arbitrary friction coefficient.

Action of the law of analogy in occasional word formation (based on the lyrics by E.A. Evtushenko and A.A. Voznesensky)

The article views analogy as one of the basic laws of language development. The action of the law of analogy is studied on the material of the occasional word creation of the poets of the 60-ies of the XX century E.A. Evtushenko and A.A. Voznesensky. The relevance of the study of occasionalisms is stipulated by insufficient knowledge of the interaction of regular and occasional word formation and the role of analogy in creating a poetic image. The main research method is the word-formation analysis of occasional units, which made it possible to analyze the derivational mechanism based on assimilation to ordinary regular models. Contextual analysis is applied, revealing the semantic side of the new lexeme. Component analysis was used to detect syntagmatic connections of occasional units. The author of the article notes the importance of the law of analogy not only in the choice of formants, but also in the procedure of the derivational process. The action of the law in the formation of occasional steps of word-formation nests is substantiated. The conditional entry of new units into the usual system and their ability to fill word-formation gaps are analyzed. In the course of the research, a lexical and word formation phenomenon of a different order was discovered a word formation paradigm, reflecting the processes of one word formation stage. The productivity of some root morphemes is traced in connection with their frequency in usage and participation in the formation of value concepts. The results of the analysis of language units confirm the importance of the law of analogy in the creation of occasionalisms and the expression of the authors assessment in general. In general, the results of the analysis of linguistic units confirm the high degree of significance of the law of analogy in the creation of occasionalisms and their use in poetic texts. The role of analogy as text formation means and a tool for constructing poetic imagery is emphasized. The conclusion is made about the systematic nature of the formation of lexical-derivational and semantic connections in the array of occasional vocabulary of E.A. Evtushenko and A.A. Voznesensky.

Reliability study of generalized exponential distribution based on inverse power law using artificial neural network with Bayesian regularization

AbstractThe investigation of lifetime reliability analysis is vital for confirming the quality of devices, equipment, electronic tube flops, and so forth. Statistical investigators have become more interested in lifetime model exploration in recent years, particularly in the last decade, without considering the issue of modeling the metrics of model reliability using artificial neural networks (ANNs). This study addresses this vacuum by discussing the multilayer ANN with Bayesian regularization modeling for reliability metrics of generalized exponential model based on inverse power law (IPL). The numerical findings of the reliability investigations and the values obtained from the ANN have been examined and analyzed carefully. The findings show that ANNs are a powerful and useful mathematical tool for analyzing the reliability of lifetime model based on IPL. Finally, a real life framework is implemented that support the theory of a research study.

LoSAC: An Efficient Local Stochastic Average Control Method for Federated Optimization

Federated optimization (FedOpt), which targets at collaboratively training a learning model across a large number of distributed clients, is vital for federated learning. The primary concerns in FedOpt can be attributed to the model divergence and communication efficiency, which significantly affect the performance. In this article, we propose a new method, i.e., LoSAC, to learn from heterogeneous distributed data more efficiently. Its key algorithmic insight is to locally update the estimate for the global full gradient after each regular local model update. Thus, LoSAC can keep clients’ information refreshed in a more compact way. In particular, we have studied the convergence result for LoSAC. Besides, the bonus of LoSAC is the ability to defend the information leakage from the recent technique Deep Leakage Gradients (DLG). Finally, experiments have verified the superiority of LoSAC comparing with state-of-the-art FedOpt algorithms. Specifically, LoSAC significantly improves communication efficiency by more than 100% on average, mitigates the model divergence problem, and equips with the defense ability against DLG.

Abstract 5400: Predicting disease progression in inoperable localized NSCLC patients using cfDNA fragmentomics assay

Abstract Background: There is an urgent clinical need to accurately predict the risk for disease progression in post-treatment NSCLC patients. However, the current ctDNA mutation profiling approaches were limited by low sensitivity, while the cfDNA fragmentomics profiling has shown an excellent capability for cancer early detection in NSCLC and therefore exhibits great potential for predicting disease progression. In this retrospective study, we aim to develop a non-invasive liquid biopsy assay utilizing cfDNA fragmentomics profiling for predicting disease progression in inoperable localized NSCLC patients. Methods: The study cohort retrospectively enrolled 44 patients diagnosed with inoperable localized NSCLC who received first-line chemoradiotherapy/radiotherapy (including 23 disease progression during follow-up). Plasma samples were collected during or post-treatments, including 39 at the fourth week of treatment (TP1), 33 and 25 at 1 (TP2) and 3 (TP3) months post-treatment, respectively. cfDNA fragmentomic profiling, generated based on target sequencing data, was used to fit Regularized Cox Regression models. For each time point, a leave-one-out cross-validation (LOOCV) was performed to evaluate the models’ predictive performances, which was subsequently compared against the ctDNA status determined by the mutation-based method. Results: Our cfDNA fragmentomics assay showed excellent performance detecting patients with a high risk for disease progression. At TP1, the high-risk patients detected by our model showed an increased risk of 3.62 times (hazard ratio [HR] = 3.62, p = 0.0026) for disease progression, compared to 3.91 times (HR = 3.91, p = 0.0022) and 4.00 times (HR = 4.00, p = 0.019) for TP2 and TP3, respectively. These fragmentomics determined HRs were higher compared to the ctDNA mutation-based results (HR = 2.08, p = 0.074; HR = 1.49, p = 0.61) at TP1 and TP3, albeit being lower at TP2 (HR = 9.47, p < 0.0001). At TP1, the predictive model reached 40% sensitivity at 92.9% specificity, which was higher than the mutation-based method (40% sensitivity at 78.6% specificity), while the combination of the two methods reached a higher sensitivity (60%). Finally, a subset of the cohort (N = 19) with data for all three time points (TP1, TP2 and TP3) was examined, and the patients were labelled as high-risk for having a positive prediction or ctDNA status at any time point. The combined methods could predict disease progression with an excellent sensitivity of 88.9% at 80% specificity. Conclusions: We developed a cfDNA fragmentomics assay for predicting disease progression in inoperable NSCLC patients. This assay showed increased predicting power during and post treatment compared to the ctDNA mutation-based method, thus illustrating a great clinical potential to guide treatment decisions in inoperable NSCLC patients. Citation Format: Yin Yang, Tao Zhang, Jingbo Wang, Jianyang Wang, Wanxiangfu Tang, Hua Bao, Haimeng Tang, Xue Wu, Yang Shao, Xin Wang, Yuqi Wu, Linfang Wu, Xin Xu, Kunpeng Xu, Jingjing Zhao, Luhua Wang, Nan Bi. Predicting disease progression in inoperable localized NSCLC patients using cfDNA fragmentomics assay. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5400.

Classification and regression-based convolutional neural network and long short-term memory configuration for bridge damage identification using long-term monitoring vibration data

Considerable attention has recently been focused on classification and regression-based convolutional neural network (CNN) and long short-term memory (LSTM) due to their excellent performance in capturing complex spatial and temporal information characteristics for structural damage identification. However, few studies have considered structural damage identification as a classification and regression problem. In addition, bridges in practical engineering are vulnerable to various environmental and vehicle loading conditions. Hence, this study proposed a new two-stage CNN–LSTM configuration for bridge damage identification using vibration data considering the influence of temperatures. First, a classification-based CNN–LSTM is designed to perform multiclass damage detection tasks, and then a regression-based CNN–LSTM is developed for damage localization and severity prediction tasks. The performance of the proposed damage identification method was evaluated through a simulation dataset of a concrete highway bridge model and a field experiment dataset of Z24-bridge (Switzerland). In addition, a set of statistical evaluation metrics such as sparse categorical cross-entropy loss, accuracy, confusion matrix, mean squared loss, mean absolute error, mean absolute percentage error, and coefficient of determination were used to compare the damage identification performance of the proposed CNN–LSTM configuration with a regular CNN model and conventional machine learning (ML) algorithms. Prediction results indicate that the proposed CNN–LSTM model outperforms the regular CNN model and conventional ML algorithms for bridge damage identification.

Fluctuation dominated phase ordering in coarse-grained depth models: domain wall structures, extreme values and coarsening

Models of particles driven by a one-dimensional fluctuating surface are known to exhibit fluctuation dominated phase ordering (FDPO), in which both the order and fluctuations appear on macroscopic scales. Highly dynamic and macroscopically broad interfacial regions, each composed of many domain walls, appear between macroscopically ordered regions and consequently the scaled correlation function violates the Porod law. We focus on two essential quantities which together quantify the unique characteristics of FDPO, namely the total number of domain walls and the length of the largest ordered domain. We present results in the context of coarse-grained depth (CD) models, both in steady state and while coarsening. Analytic arguments supported by numerical simulations show that even though domain wall number fluctuations are very strong, the associated variance remains constant in time during coarsening. Further, the length of the largest cluster grows as a power law with significant multiplicative logarithms which involve both the time and system size. In addition, we identify corrections to the leading power law scaling in several quantities in the coarsening regime. We also study a generalisation of the CD model in which the domain wall density is controlled by a fugacity and show that it maps on to the truncated inverse distance squared Ising model. The generalised model shows a mixed order phase transition, with the regular CD model (which exhibits FDPO) corresponding to the critical point.

Stabilization of Equiatomic Solutions Due to High-Entropy Effect

The stability of solid-solution phases in FCC and BCC lattices was examined in multi-component alloys based on the CALPHAD technique using the compound energy formalism and regular solution model. From the thermodynamic calculations, it was found in ternary systems that the single solid-solution phase became stable around the equiatomic composition where the configurational entropy was the largest value. The transition temperature from the disordered phase to ordered phase(s) or miscibility gap(s) decreased with the increasing number of elements in the system. The order-disorder transition temperature on the FCC lattice was affected by the end member of the ordered phases existing at the equiatomic composition, whereas it was not significant for the order-disorder transition in the BCC lattice. The single solid-solution phase region at equiatomic compositions was affected by variations in the interaction parameters. In multi-component systems, the variations were averaged with increasing the number of elements in the system. This suggests that high-entropy alloys can afford a variety of elements. This study shows that the disordered state can be formed in multicomponent systems around the equiatomic composition and suggests clearly that due to the high-entropy effect, the solution phases are stabilized.

Data-Driven State of Health Estimation for Lithium-Ion Batteries Based on Universal Feature Selection

A simple yet effective health indicator (HI)-based data-driven model forecasting the state of health (SOH) of lithium-ion batteries (LIBs) and thus enabling their efficient management is developed. Five HIs with high physical significance and predictive power extracted from voltage, current, and temperature profiles are used as model inputs. The generalizability and robustness of the proposed ridge regression–based linear regularization model are assessed using three NASA datasets containing information on the behavior of batteries over a wide range of temperatures and discharge rates. The maximum mean absolute error, maximum root-mean-square error, and maximum mean absolute percentage error of the SOH for the three groups of batteries are determined as 0.7%, 0.86%, and 2.1%, respectively. Thus, the developed model exhibits high accuracy in estimating the SOH of LIBs under multiworking conditions and is sufficiently robust to be applicable to low-quality datasets obtained under other conditions.

A fraction function regularization model for the structural damage identification

The conventional model updating based on sensitivity analysis generally employs l1-norm regularizer to characterize the sparsity of the structural damage. However, the l1-norm regularizer inevitably excessively penalizes the larger components in the damage parameter, which certainly causes the extra estimation bias of the damage parameter and reduces the damage identification accuracy. A fraction function regularizer not only well characterizes the sparsity, but also overcomes the excessive penalty drawback of the l1-norm regularizer. Based on this, a fraction function regularization model is proposed to improve the damage identification accuracy. Numerical and experimental results illustrate that the damage identification accuracy of the proposed model is averagely improved 4.96% and 3.68% than that of the l1 regularization one, the iteratively reweighted l1 regularization one and the elastic net one, respectively.

Two Strategies for Avoiding Overfitting Long-Term Forecasting Models: Downsampling Predictor Fields and Shrinking Coefficients

Long-term hydrological forecasting based on sea surface temperature (SST) fields faces the large p and small n problem, i.e., too many potential predictors and a limited number of samples. Considering the selection of predictors will also enhance the complexity of models and lead to overfitting, in this study, two strategies are used for building forecasting models for long-term streamflow forecasting. The first strategy is to downsample the SST field and optimize its spatial resolution; the second is to shrink model coefficients based on L1 regularization. We build models based on the downsampled SST fields with different spatial resolutions. It is found that the model based on a proper spatial resolution always performs better than the model based on the raw SST field. This result suggests that it is better to treat the spatial resolution of the predictor field as a hyperparameter, which is similar to hyperparameters for controlling the complexity of many machine learning models. For applying the second strategy, L1 norm regularization models, including (1) least absolute selection and shrinkage operator (LASSO), (2) relaxed LASSO, and (3) two-step approach of LASSO and ordinary least squares regression (LASSO+OLS) are explored. We have found that the relaxed LASSO model always performs better than the ordinary LASSO, indicating that relaxed LASSO is a better shrinkage approach. Furthermore, the skills of the presented models are compared with the stepwise regression, and the lower skill of the stepwise regression based on SST fields with a high spatial resolution suggests that one should not select predictors based on fields with high resolutions considering the limited number of samples.