Hybrid Modeling Approach Research Articles

Uncovering the generic and alloy-specific governing parameters of deformation-induced martensitic transformation in austenitic steel

In this work, a hybrid modeling approach, combining machine learning (ML) and computational thermodynamics, has been applied to predict deformation-induced martensitic transformation (DIMT) and explore the generic and alloy-specific parameters governing DIMT in austenitic steels. The DIMT model was established based on the ensemble ML algorithms and a comprehensive set of physical variables. The developed model is highly generalizable as validated on unseen alloys. The generic governing parameters of DIMT are in good agreement with previous studies in the literature. However, the evaluated alloy-specific governing parameters reveal large differences between grades, e.g., 204 series of austenitic stainless steels has a quite balanced correlation between strain, stress, temperature, and DIMT, while the 301 series has much stronger correlation between stress and DIMT. The findings in the current study emphasize the importance that a general DIMT model for steels should include both stress and strain, as well as other governing parameters, since DIMT can be both stress-assisted and strain-induced transformation, and often the effect of applied mechanical driving force and the formation of new nucleation sites interact.Graphical abstract

Physics-based and data-driven hybrid modeling in manufacturing: a review

ABSTRACT Manufacturing, an industry set in the physical world, is undergoing its digital transformation, also known as the fourth industrial revolution. Sensors, connectivity, and platforms provide unprecedented access to quantify the quality and diversity of manufacturing data. Progress in data-driven modeling is exponential across all industries. This leads to the question of how physics-based and data-driven modeling can be utilized in a hybrid modeling approach to advance our understanding of processes, materials, and systems in manufacturing. In this review, we focus on discrete manufacturing based on the understanding that hybrid modeling is more mature in process manufacturing. This paper aims to provide an overview of projects where hybrid modeling was used in manufacturing and introduce various ways of composing hybrid models. We provide examples highlighting the implementation of models, structure and expand on metrics to test and validate hybrid models, discuss challenges, and future research directions of hybrid modeling in manufacturing.

Remaining useful life prediction of lithium battery based on CEEMD-SE-IPSO-LSSVM hybrid model

Abstract In order to prevent accidents caused by battery aging, accurately predicting the remaining useful life (RUL) is a critical and highly challenging task in battery management systems. This article describes a lithium-ion battery RUL prediction method based on a hybrid model of CEEMD-SE-IPSO-LSSVM. This method integrates various technologies and algorithms, enhancing the accuracy and practicality of predictions. Initially, the complete ensemble empirical mode decomposition (CEEMD) is utilized to decompose the raw data into multiple intrinsic mode functions, aiding in denoising and feature extraction. Subsequently, the sample entropy (SE) is used to assess the complexity and irregularity of the data, merging intrinsic mode function components with similar SE values into a new component. Building upon this, the advanced iterative particle swarm optimization (IPSO) algorithm refines the parameters of the least squares support vector machine (LSSVM) model, improving the predictive performance of the model. Finally, through iterative training and refinement of the LSSVM model, accurate prediction of the remaining life of lithium-ion batteries is achieved. This hybrid model approach integrates data processing, feature extraction, and model refinement, resulting in a significant improvement over the baseline model with a 69.5% increase in mean absolute percentage error and a 49.4% decrease in root mean squared error, providing a robust solution for predicting the remaining life of lithium-ion batteries.

Assessing the speculative dynamics and determinants of residential apartment rentals in Mogadishu, Somalia: A hybrid modeling approach

There has been relatively rapid growth in Somalia's real estate sector recently, resulting in property market booms in major cities of the country. Despite this sector's importance, very little scientific research has been conducted. Thus, this study aims to model the rental value of residential apartments in Mogadishu, Somalia. A hybrid modeling approach is utilized, in which a hedonic regression model was used in the first phase and an Artificial Neural Network (ANN) in the second. After analysis, the study established that an apartment's age, size, value, view, number of toilets, air quality, proximity to CBD, proximity to a university and probability of explosion are key factors determining apartment rental rates in Mogadishu, while floor level, parking space, and school proximity do not have a significant impact on apartment rental rates. Because of speculation, there is a higher likelihood of overvaluation than undervaluation in the housing market, which has profound policy and practical implications. Rather than being driven by real demand, Mogadishu's real estate market is driven by speculation, as overvaluation signals are more evident than undervaluation signals. As one of the world's poorest nations, where 70% of the population lives below the international poverty line, and about 80% of jobs are in the informal sector, extreme speculative activities adversely affect Somalia's economic quality and the living conditions of its citizens. Somali citizens need government policies that facilitate the accessibility, affordability, and adequate availability of decent housing. The study recommends protecting Somalia's real estate sector to attract more investors and boost the country's post-conflict development initiatives. A vital contribution of the study is that it is the first to examine the rental value of residential apartments in Mogadishu systematically. This study contributes significantly to housing economics and real estate development literature.

Correction of Thermal Errors in Machine Tools by a Hybrid Model Approach

Thermally induced position errors are one of the main error sources on the workpiece caused by the behavior of the machine tool. In today’s industrial environment, the correction of thermal errors is usually based on simple regression approaches, where the characteristic diagrams for correction are generated experimentally. The performance of these approaches is only valid for the corresponding load regimes, which often results in insufficient correction quality in practical applications. Consequently, there is only a limited benefit or even a deterioration in machine behavior if the correction characteristic is based on an inapplicable load case compared to the initial experiment. Simulation-generated characteristic diagrams using finite element models solve this disadvantage, but do not answer the question about the choice of the right characteristic matching the current load situation, and, in addition, calculate very slowly. Structural model-based correction using reduced models, on the other hand, calculates quickly, but requires a high modeling effort for accurate correction. The approach, presented in this contribution, combines simulation-generated characteristic diagrams and a structural model-based decision algorithm for a new hybrid model in order to select the appropriate characteristic diagram for the present load situation in the control system. This paper presents the simulative characteristic diagram generation by a finite element model validated by experiments in a climate chamber and a validated structural model including the concept for the decision algorithm.

A Framework for Analyzing System Loadability With Multiple VSCs Using a Hybrid Model

Due to the increasing number of power electronics-based devices in modern power systems, there are concerns about the impact of higher-frequency interactions on system stability. The typically-used algebraic representation of networks fails to capture these newly emergent issues, but dynamically modelling the entire network with differential equations will lead to a prohibitively long simulation time. This paper proposes a computationally-efficient analysis framework to determine the system loadability in power electronics-rich, large networks with the consideration of multiple types of stabilities. The framework developed in this paper identifies the critical network elements based on the eigenvalue sensitivity in order to exploit a hybrid modelling approach. Within the hybrid model, only the critical network portion is modelled dynamically with the rest of the network represented by algebraic equations. Bifurcation theory is used to simultaneously analyze both small-disturbance rotor angle and small-disturbance voltage stability. The results obtained show that the high-frequency interactions are accurately captured with the proposed framework. Additionally, applying this methodology results in a small dimension for the system matrix and a reduction in the computational burden. Two test networks, a three-bus system and the IEEE 39-bus system, are used to illustrate and verify the analysis framework.

High-Precision Dynamic Control of Soft Robots With the Physics-Learning Hybrid Modeling Approach

High-Precision Dynamic Control of Soft Robots With the Physics-Learning Hybrid Modeling Approach

3D hybrid modeling approach combining the finite and discrete element methods: Validation based on masonry shear wall tests

The masonry structures represent a remarkable portion of the world’s built heritage. The preservation of these buildings requires a multidisciplinary diagnostic approach that includes a structural assessment to estimate their stability. However, this task remains difficult because of the high nonlinearity behavior of the masonry material and the complexity of the geometries of this type of building. For this reason, several robust modeling strategies have been developed. Among these strategies, block-based models (BBM) based on the discrete element method and more specifically the NSCD “Non Smooth Contact Dynamics” allows to take into account the actual geometrical texture of masonry. However, in this type of approach, blocks are generally rigid or elastic and nonlinearity appears only at the interfaces between blocks governed by cohesive frictional laws. In this study, a new hybrid approach is proposed. It consists of integrating a model combining (i) a frictional cohesive zone model “FCZM” describing the block-mortar interface and (ii) a quasi-brittle damage model associated with the block while (iii) the elasticity of mortar joints is explicitly taken into account in the FCZM model by adding a linear spring in both normal and tangential directions. Thus, avoiding the use of extended blocks and corresponding homogeneous properties. The validation of this numerical approach is carried out on the basis of an experimental campaign on masonry shear wall test subjected to compression.

Enhancing Early Detection of Blood Disorders through A Novel Hybrid Modeling Approach

Blood disorders are such conditions that impact the blood’s ability to function correctly. There is a range of different symptoms depending on the type. There are several different types of blood disorders such as Leukemia, chronic myelocytic leukemia, lymphoma, myelofibrosis, polycythemia, thrombocytopenia, anemia, and leukocytosis. Some resolve completely with therapy or do not cause symptoms and do not affect overall lifespan. Some are chronic and lifelong but do not affect how an individual lives. Other blood disorders, like sickle cell disease and blood cancers, can be even fatal. There needs to be a capture of hidden information in the medical data for detecting diseases in the early stages. This paper presents a novel hybrid modeling strategy that makes use of the synergy between two methods with histogram-based gradient boosting classifier tree and random subspace. It should be emphasized that the combination of these two models is being employed in this study for the first time. We present this novel model built for the assessment of blood diseases. The results show that the proposed model can predict the tumor of blood disease better than the other classifiers.

A hybrid modelling approach for detecting seasonal variations in inland Green-Blue Ecosystems

Deforestation, environmental pollution, and the overexploitation of resources, in addition to the Earth's natural cycles, are scaling up the impacts of climate change in the provision of Ecosystem Services (ES). Green-Blue Ecosystems (GBE) are impacted by climatic conditions, topography, and water presence. Data-driven modelling techniques may effectively capture the effects of seasonal variations while modelling natural ecosystems. This research proposes a hybrid modelling approach that combines Deep Learning and traditional Machine Learning, Sensitivity Analysis and Feature Importance Evaluation (FIE) to investigate seasonality effects on mapping GBE. The models, built using satellite imagery from the Spring and Summer seasons of the Mediterranean climate zone, included spectral indices, topography (DEM), and groundwater depth (GD). The model that best suited the analysis was selected using sensitivity tests and hyperparameter optimization. The study shows that land cover classes of transitional woodland shrubs, inland marshes, cultivated land parcels, and watercourses are better classified in the Spring, with an accuracy of 0.814. FIE indicates that spectral indices are the most important predictors for detecting green ecosystems in both seasons. Additionally, DEM and GD are the most relevant predictors to classify watercourses in the Summer. An analytical examination of the input data and hyperparameter settings facilitates understanding of models' behaviour while improving models' prediction. This research provides an advanced understanding of the effects of seasonal variations on the status of GBE and enhances understanding of modelling ES in areas with a growing need for changes in land use and high water supply demand.

Application of response surface Methodology coupled with Artificial Neural network and genetic algorithm to model and optimize symbiotic interactions between Chlorella vulgaris and Stutzerimonas stutzeri strain J3BG for chlorophyll accumulation

Research on microalgae has surged due to its diverse biotechnological applications and capacity for accumulating bioactive compounds. Despite considerable advancements, microalgal cultivation remains costly, prompting efforts to reduce expenses while enhancing productivity. This study proposes a cost-effective approach through the coculture of microalgae and bacteria, exploiting mutualistic interactions. An engineered consortium of Chlorella vulgaris and Stutzerimonas stutzeri strain J3BG demonstrated biofilm-like arrangements, indicative of direct cell-to-cell interactions and metabolite exchange. Strain J3BG's enzymatic characterization revealed amylase, lipase, and protease production, sustaining mutual growth. Employing Response Surface Methodology (RSM), Artificial Neural Network (ANN), and Genetic Algorithm (GA) in a hybrid modeling approach resulted in a 2.1-fold increase in chlorophyll production. Optimized conditions included a NaNO3 concentration of 128.52 mg/l, a 1:2 (Algae:Bacteria) ratio, a 6-day cultivation period, and a pH of 5.4, yielding 10.92 ± 0.88 mg/l chlorophyll concentration.

Physically Based Thermal Infrared Snow/Ice Surface Emissivity for Fast Radiative Transfer Models

Accurate thermal infrared (TIR) fast-forward models are critical for weather forecasting via numerical weather prediction (NWP) satellite radiance assimilation and operational environmental data record (EDR) retrieval algorithms. The thermodynamic and compositional data about the surface and lower troposphere are derived from semi-transparent TIR window bands (i.e., surface-sensitive channels) that can span into the far-infrared (FIR) region under dry polar conditions. To model the satellite observed radiance within these bands, an accurate a priori emissivity is necessary for the surface in question, usually provided in the form of a physical or empirical model. To address the needs of hyperspectral TIR satellite radiance assimilation, this paper discusses the research, development, and preliminary validation of a physically based snow/ice emissivity model designed for practical implementation within operational fast-forward models such as the U.S. National Oceanic and Atmospheric Administration (NOAA) Community Radiative Transfer Model (CRTM). To accommodate the range of snow grain sizes, a hybrid modeling approach is adopted, combining a layer scattering model based on the Mie theory (viz., the Wiscombe–Warren 1980 snow albedo model, its complete derivation provided in the Appendices) with a specular facet model. The Mie-scattering model is valid for the smallest snow grain sizes typical of fresh snow and frost, whereas the specular facet model is better suited for the larger sizes and welded snow surfaces typical of aged snow. Comparisons of the model against the previously published spectral emissivity measurements show reasonable agreement across zenith observing angles and snow grain sizes, and preliminary observing system experiments (OSEs) have revealed notable improvements in snow/ice surface window channel calculations versus hyperspectral TIR satellite observations within the NOAA NWP radiance assimilation system.

Hourly rainfall-runoff modelling by combining the conceptual model with machine learning models in mostly karst Ljubljanica River catchment in Slovenia

Hydrological modelling, essential for water resources management, can be very complex in karst catchments with different climatic and geologic characteristics. In this study, three combined conceptual models incorporating the snow module with machine learning models were used for hourly rainfall-runoff modelling in the mostly karst Ljubljanica River catchment, Slovenia. Wavelet-based Extreme Learning Machine (WELM) and Wavelet-based Regression Tree (WRT) machine learning models were integrated into the conceptual CemaNeige Génie Rural à 4 paramètres Horaires (CemaNeige GR4H). In this regard, the performance of the hybrid models was compared with stand-alone conceptual and machine learning models. The stand-alone WELM and WRT models using only meteorological variables performed poorly for hourly runoff forecasting. The CemaNeige GR4H model as stand-alone model yielded good performance; however, it overestimated low flows. The hybrid CemaNeige GR4H-WELM and CemaNeige-WRT models provided better simulation results than the stand-alone models, especially regarding the extreme flows. The results of the study demonstrated that using different variables from the conceptual model, including the snow module, in the machine learning models as input data can significantly affect the performance of rainfall-runoff modelling. The hybrid modelling approach can potentially improve runoff simulation performance in karst catchments with diversified geological formations where the rainfall-runoff process is more complex.

A hybrid modeling strategy for training data generation in machine learning-based structural health monitoring

Concerning the cost- and resource-saving maintenance of assembly products, it is vital to detect any potential malfunctions, defects or structural damage at the earliest-possible stage. For this reason, considerable efforts are being put into the development of Structural Health Monitoring, a field encompassing different approaches to damage identification and capable of preventing defects and even failure. Structural Health Monitoring is often supported by machine learning, a tool for rapid and effective damage identification that can recognize patterns or changes in the data received from the structure. Despite the advances machine learning has made in recent years, obtaining a suitable data set for the efficient training of machine learning algorithms within Structural Health Monitoring remains a challenge. Currently, the data are usually obtained experimentally, with numerical or analytical models. However, the experimental approach can often be time consuming, while the reliability of numerically obtained data relies heavily on the accuracy of the numerical models in capturing the true behavior of the structure. Analytical models may be constrained by the complexity of the observed object. In this paper an alternative approach based on an experimental–numerical (i.e., hybrid) modeling approach is proposed to build a training set for Structural Health Monitoring. Frequency Based Substructuring is utilized to determine the response model of the assembled system based on the properties of its components as well as to mix experimental and numerical models, while leveraging the advantages of each. This makes it possible to generate the samples of the training set in the form of hybrid models of the structure of interest, exhibiting the realistic properties of a physical structure, with a reasonable measurement effort. Here, the approach is demonstrated for the process of joint-damage identification.

Spatially-explicit absolute life cycle assessment by multi-regional hybrid modeling: Computational framework

Human activities, environmental impacts, and the provision of ecosystem services vary by region. The existing regional environmental sustainability assessment approaches addressed such regional heterogeneity of inventory data at a single scale (i.e., either process or economy scale). However, depending on regional data availability and the assessment scope, a multiscale modeling approach that integrates existing data and models at various scales needs to be employed. This work introduces a general computational framework for spatially-explicit environmental sustainability assessment based on a multi-regional hybrid modeling approach. The developed framework accounts for regional activities at multiple scales by integrating the existing regional approaches at each scale. Also, the framework integrates the supply of ecosystem services, which vary by region as well, at multiple scales to perform spatially-explicit absolute environmental sustainability assessment. The proposed framework is demonstrated using an illustrative example of corn production in two regions. The results are discussed in comparison to the conventional approaches.

An investigation of the effects of ironing parameters on the surface and compression properties of material extrusion components utilizing a hybrid-modeling experimental approach

Thermoplastic filament Material Extrusion (ME) is one of the most expansive 3D printing processes. Owed to the ME process’s simplicity, low cost of materials, popularity, and flexibility, considerable attention has been devoted to manufacturing specific parts in industries manipulating the polylactic acid (PLA) polymer, such as automotive and biomedical. This work aims to experimentally investigate the material flow and temperature for different layer heights on the surface texture parameters and compression strength of a tailored PLA hexagonal prismatic part. An experimental approach based on L9 Taguchi’s array and residual analysis (ANOVA) was employed to clarify the parameters’ effects and trends regarding the response variables. The analysis of means showed that the material flow and layer height are critical variables in defining ME parts’ roughness and compression. Based on ANOVA and mean absolute percentage errors (MAPE) results, additive models (ADMO) were used to predict all combinatorial response values. Then, the experimental and the ADMO values feed as trained data for developing a feed-forward back-propagation neural network (FFBP-NN). Three independent experiments confirmed the validity of the proposed methodology resulting in reasonable accuracy of all the performance metrics, making the proposed hybrid-modeling approach adequate for process multi-parameter multi-objective optimization 3D printing cases.

Formulating Mass-loss Rates for Sun-like Stars: A Hybrid Model Approach

We observe an enhanced stellar wind mass-loss rate from low-mass stars exhibiting higher X-ray flux. This trend, however, does not align with the Sun, where no evident correlation between X-ray flux and mass-loss rate is present. To reconcile these observations, we propose a hybrid model for the stellar wind from solar-type stars, incorporating both Alfvén wave dynamics and flux emergence-driven interchange reconnection, an increasingly studied concept guided by the latest heliospheric observations. For establishing a mass-loss rate scaling law, we perform a series of magnetohydrodynamic simulations across varied magnetic activities. Through a parameter survey concerning the surface (unsigned) magnetic flux (Φsurf) and the open-to-surface magnetic flux ratio (ξ open = Φopen/Φsurf), we derive a scaling law of the mass-loss rate given by , where , , and . By comparing cases with and without flux emergence, we find that the increase in the mass-loss rate with the surface magnetic flux can be attributed to the influence of flux emergence. Our scaling law demonstrates an agreement with solar wind observations spanning 40 yr, exhibiting superior performance when compared to X-ray-based estimations. Our findings suggest that flux emergence may play a significant role in the stellar winds of low-mass stars, particularly those originating from magnetically active stars.

A Deep Learning Approach Based on Physical Constraints for Predicting Soil Moisture in Unsaturated Zones

AbstractWater transport in the unsaturated zone is an important part of the hydrological cycle and is the link between the atmosphere‒soil‐groundwater for material and energy transport. The accurate prediction of soil moisture (SM) is essential for the rational exploitation of water resources. Data‐driven deep learning methods are widely used in many fields; however, the lack of physical mechanisms limits their application in hydrological fields, especially for SM prediction in unsaturated zones. To solve this problem, this study proposes a new deep learning method that introduces the water balance principle, Richard's equation, and SM boundary conditions as constraints to construct the new loss function that guides the training process of deep learning, called physics‐informed deep learning (PIDL). In tests consisting of a large number of data sets acquired from in situ observation sites in the field, PIDL exhibits higher accuracy than ordinary deep learning (long short‐term memory) and physical models, with 51.03% and 53.46% reduction in root mean square error of SM prediction, respectively. PIDL performance significantly improved in predicting scenarios that are difficult for ordinary deep learning to handle, such as sparse data sets, extreme values, and mutated values. In addition, PIDL maintains high accuracy over a longer prediction period. The addition of physical mechanisms allows deep learning to mine patterns not only from the data itself but also from a priori physical theoretical knowledge for guidance, and this hybrid modeling approach can also be generalized to prediction problems in other hydrological domains.

Redefining hybrid flow shop group scheduling: Unveiling a novel hybrid modeling paradigm and assessing 48 MILP and CP models

A hybrid flow shop group scheduling problem (HFGSP) involves two distinct sub-problems, the arrangement of groups and the configuration of jobs within each group. Constructing its mathematical model based on the classification of these sub-problems can help better understand the problem's characteristics. However, the existing literature fails to establish MILP models of HFGSP based on the interrelation between each subproblem and the respective process constraints satisfied by each subproblem. To address this gap, our study first categorizes all the constraints in the HFGSP according to several factors: group arrangement, job arrangement within the group, process requirements between groups, process requirements between jobs from the same group, process requirements between adjacent stages, and the relationship between the decision variables of completion time and the objective constraint. Following this classification, we construct 48 available MILP models of HFGSP. Additionally, we also propose an efficient constraint programming (CP) model of HFGSP. Numerous experiments are conducted to verify the correctness and performance of all 48 MILP models across different test instances, analyze the complexities of all models, and excavate their intrinsic characteristics. Through our evaluation for the 48 models, we observe that models 24 and 48 exhibit superior performance. This highlights the effectiveness of the hybrid modeling approach, which synergistically combines sequence-based modeling for groups and position-based adjacent modeling for jobs within each group. The utilization of this hybrid modeling approach enables the construction of high-quality MILP models for addressing the HFGSP problem. Our intention is to bridge the gap between MILP modeling of HFGSP and problem-specific algorithms, thereby providing a comprehensive understanding of the problem and offering potential solutions.

Integrating APSIM model with machine learning to predict wheat yield spatial distribution

AbstractTraditional simulation models are often point based; thus, more research is needed to emphasize spatial simulation, providing decision‐makers with fast recommendations. Combining machine learning algorithms with spatial process‐based models could be considered an appropriate solution. We created a spatial model in R (APSIMx_R) to generate fine‐resolution data from coarse‐resolution data, which is typically available at the regional level. The APSIM crop model outputs were then deployed to train and test the artificial neural network, creating a hybrid modeling approach for robust spatial simulations. The APSIMx_R package facilitates preparing the required model inputs, executes the prediction, processes, and analyzes the APSIM crop model outputs. This note demonstrates the use of a new approach for creating reproducible crop modeling workflows with the spatial APSIM next‐generation model and machine learning algorithms. The tool was deployed for spatial and temporal simulation of potential wheat yield under different nitrogen rates and various wheat cultivars. The spatial APSIMx_R was validated by comparing the simulated yield at 100 kg N ha−1 to the analogues' actual yield at the same grid points, which showed good agreement (d = 0.89) between the spatially predicted and actual yield. The hybrid approach increased such precision, resulting in higher agreement (d = 0.95) with actual yield. When the interaction between cultivars and nitrogen levels was considered, it was found that the novel cultivar Sakha95 is nitrogen voracious, exhibiting a larger drop in yield (65%) under minimal nitrogen treatment (0 kg N ha−1) relative to the potential yield.