Multi-domain Model Research Articles

Code Generation Approach Supporting Complex System Modeling based on Graph Pattern Matching

Code generation is an effective way to drive the complex system development in model-based systems engineering. Currently, different code generators are developed for different modeling languages to deal with the development of systems with multi-domain. There are a lack of unified code generation approaches for multi-domain heterogeneous models. In addition, existing methods lack the ability to flexibly query and transform complex model structures to the target code, resulting in poor transformation efficiency. To solve the above problems, this paper proposes a unified approach which supports the code generation of heterogeneous models with complex model structure. First, The KARMA language based on GOPPRR-E meta-modeling approach is used for the unified formalism of models built by different modeling languages. Second, the code generation approach based on graph pattern matching is proposed to realize the query and transformation of complex model structures. Then, the syntax for code generation is integrated into KARMA and a compiler for code generation is developed. Finally, a case of unmanned vehicle system is taken to validate the effectiveness of the proposed approach.

Supervisory Real-Time Multidomain Modeling and Hardware Emulation of Fuel-Cell Hybrid Electric Bus Behavioral Transients

Real-time device-level multi-domain emulation can provide an accurate insight into behavioral transients of the hydrogen fuel-cell hybrid electric bus (HEB). However, the conventional electromagnetic transient (EMT) simulation suffers from the computation burden caused by the complex multi-domain subsystems. This paper develops a hybrid recurrent neural network (RNN) and EMT method for device-level multi-domain emulation for fuel-cell and battery HEB. Two recurrent neural networks (RNN) are designed and trained to create device-level models of permanent magnet synchronous motor (PMSM) and the modular multilevel converter (MMC), respectively. The IGBTs’ behavioral transients and thermal performance are integrated into the RNN-based MMC model. Moreover, the EMT models represent the energy behaviour of onboard fuel-cell stacks and battery stacks. The proposed multi-domain hybrid models are implemented on the Xilinx Versal™ adaptive compute acceleration platform (ACAP), where multiple AI engines and the programmable logic deal with the RNN and EMT models, respectively. The real-time hardware emulation is carried out at the time-step of 0.1 μs for device-level transients. The results show that the hybrid model has 96.3% accuracy; furthermore, it significantly reduces the HEB emulation time compared to conventional EMT methods.

Cognitive reserve. A proposal for a new conceptual hypothesis

Currently, there is a lack of consensus on the definition of the concept of cognitive reserve, the elements that make it up and the relationship between them, despite the notable increase in the number of studies conducted on the subject. A multi-domain conceptual model with two factors is proposed: general reserve and domain-specific reserve. The domain-specific reserve, in turn, would be divided into four areas: cognitive, behavioural, emotional and social. The proposed model seems to provide a more comprehensive explanation of the construct, with new elements of study to be taken into consideration. This proposal aims to introduce a new perspective into the study of cognitive reserve and to open up new lines of clinical and academic research with the aim of demonstrating its feasibility.

The Impact of Semi-Supervised Learning on the Performance of Intelligent Chatbot System

Artificial intelligent based dialog systems are getting attention from both business and academic communities. The key parts for such intelligent chatbot systems are domain classification, intent detection, and named entity recognition. Various supervised, unsupervised, and hybrid approaches are used to detect each field. Such intelligent systems, also called natural language understanding systems analyze user requests in sequential order: domain classification, intent, and entity recognition based on the semantic rules of the classified domain. This sequential approach propagates the downstream error; i.e., if the domain classification model fails to classify the domain, intent and entity recognition fail. Furthermore, training such intelligent system necessitates a large number of user-annotated datasets for each domain. This study proposes a single joint predictive deep neural network framework based on long short-term memory using only a small user-annotated dataset to address these issues. It investigates value added by incorporating unlabeled data from user chatting logs into multi-domain spoken language understanding systems. Systematic experimental analysis of the proposed joint frameworks, along with the semi-supervised multi-domain model, using open-source annotated and unannotated utterances shows robust improvement in the predictive performance of the proposed multi-domain intelligent chatbot over a base joint model and joint model based on adversarial learning.

A Study of Thermal and Electrical Multi-Domain Model of Aluminum Electrolytic Capacitors (First Report)

A Study of Thermal and Electrical Multi-Domain Model of Aluminum Electrolytic Capacitors (First Report)

Finite volume element methods for a multi-dimensional fracture model

This study concerns the Darcy flow problem in a two-dimensional fractured porous domain, in which the fracture is regarded as a one-dimensional interface, interacting with the surrounding media. In this paper, a finite volume element method (FVEM) is first proposed for the multi-dimensional fracture model, and error estimates for the pressure with optimal convergence are discussed. On this basis, a two-grid FVEM is developed for decoupling the multi-domain fracture model by a coarse grid approximation to the interface coupling conditions, and theoretical analysis demonstrates that approximation accuracy does not deteriorate under the two-grid decoupling technique. Finally, numerical experiments for FVEM and two-grid FVEM are presented to confirm the accuracy of theoretical analysis.

Multi-Domain Dynamic Modelling of a Low-Cost Upper Limb Rehabilitation Robot

Tracking patient progress through a course of robotic tele-rehabilitation requires constant position data logging and comparison, alongside periodic testing with no powered assistance. The test data must be compared with previous test attempts and an ideal baseline, for which a good understanding of the dynamics of the robot is required. The traditional dynamic modelling techniques for serial chain robotics, which involve forming and solving equations of motion, do not adequately describe the multi-domain phenomena that affect the movement of the rehabilitation robot. In this study, a multi-domain dynamic model for an upper limb rehabilitation robot is described. The model, built using a combination of MATLAB, SimScape, and SimScape Multibody, comprises the mechanical electro-mechanical and control domains. The performance of the model was validated against the performance of the robot when unloaded and when loaded with a human arm proxy. It is shown that this combination of software is appropriate for building a dynamic model of the robot and provides advantages over the traditional modelling approach. It is demonstrated that the responses of the model match the responses of the robot with acceptable accuracy, though the inability to model backlash was a limitation.

High-frequency voltage-driven vibrations in dielectric elastomer membranes

This article deals with modelling and experimental characterisation of the continuum dynamic response of dielectric elastomer actuators (DEAs) subject to high-frequency voltage excitation.DEAs are capable of large deformations in response to an electrostatic stimulus, and they show large operating bandwidths of up to a few kilohertz. Although DEA systems normally make use of simple deformation patterns and a well-defined main actuation mode, they show complex structural dynamics and mode shapes if subject to high frequency voltage inputs. Taking advantage of these complex structural dynamics potentially allows developing multi-function actuators, audio devices and vibration isolators capable to perform different tasks using different vibration regimes.We first present a multi-domain model for the structural dynamics of DEAs, accounting for the contribution of the air pressure loads generated the DEA membrane vibrations, which play a relevant role in the DEA dynamics.We then present an extensive experimental characterisation of DEA samples’ structural dynamics based on laser Doppler vibrometer measurements. In particular, we measure the complex structural mode shapes and the velocity spectra generated through a broadband excitation in correspondence of a wide set of different design and control parameters (namely, the DEA geometric layout, the mechanical-preload and the applied voltage bias).Based on the experimental results, we validate the proposed continuum model, and we demonstrate that the forced response of the DEA can be efficiently described using a lumped-parameter computationally efficient reduced reformulation.

Equivalent Circuit Modelling of Li-S Batteries

The presented article describes a Multi-Scale Multi-Domain model of a Li-S battery using an equivalent circuit model consisting of R-C pairs. The parameters of the equivalent circuit model were obtained by fitting experimentally measured results from electrochemical impedance spectroscopy at various states of charge. Simulation were carried out for discharging as well as for charging of the battery. The presented model is able to capture characteristic behaviour of Li-S batteries.

Evaluating dual-domain models for upscaling multicomponent reactive transport in mine waste rock

Reactive transport models have proven abilities to simulate the quantity and quality of drainage from mine waste rock. Tracer experiments indicate the presence of fast and slow flow regimes in many heterogeneous waste-rock piles. Although multidomain models have been developed specifically for systems with such distinctive hydrodynamics, there have been limited applications of multidomain reactive transport models to simulate composite drainage chemistries from waste-rock piles to date. This work evaluated the ability of dual-domain multicomponent reactive transport models (DDMRTMs) to reproduce breakthrough curves of conservative (chloride) and reactive (molybdenum) solutes observed at a well-characterized experimental waste-rock pile at the Antamina Mine, Peru. We found that the DDMRTM simulations quantitatively matched eight-year-long records of conservative transport through the waste-rock pile when parameterized mainly with field-measured properties obtained from the site and limited calibration. The DDMRTM model also provided a reasonable match to field observations of the reactive solute. The limited calibrated parameters are physically realistic, corroborating the ability of these multidomain models to reproduce the complex reactive-transport processes governing polluted rock drainage from large-scale waste-rock piles.

Heat Transfer Model of Pneumatic End-Position Cylinder Cushioning

In this paper a thermal model of a pneumatic cylinder with an integrated pneumatic end cushioning is presented. Being a part of a multidomain model presented in former research, this model is needed to simulate and analyse the thermodynamic processes in the pneumatic end cushioning and to elaborate a novel design strategy for damping systems with a higher capability on kinetic energy absorption and robust performance under fluctuating operating conditions. For this purpose, a proper heat exchange model is inevitable to calculate the pressure in the cushioning volume and consequently the deceleration of the load. An approach of splitting the complex geometry of cylinder into simple geometries, such as plain or cylindrical surfaces, is used in this study for a fast computation of convective heat flow rates. To validate this approach, the simulation results were compared with the measurements, carried out at different supply pressures, piston speeds and end cushioning throttle openings. The model will be used further for sensitivity analysis and robust optimisation of the cushioning system design.

Three-Dimensional Thermal Simulations of 18650 Lithium-Ion Batteries Cooled by Different Schemes under High Rate Discharging and External Shorting Conditions

In this work, three-dimensional thermal simulations of single 18650 lithium-ion battery cell and 75 V lithium-ion battery pack composed of 21 18650 battery cells are performed based on a multi-scale multi-domain (MSMD) battery modeling approach. Different cooling approaches’ effects on 18650 lithium-ion battery and battery pack thermal management under fast discharging and external shorting conditions are investigated and compared. It is found that for the natural convection, forced air cooling, and/or mini-channel liquid cooling approaches, the temperature of battery cell easily exceeds 40 °C under 3C rate discharging condition. While under external shorting condition, the temperature of cell rises sharply and reaches the 80 °C in a short period of time, which can trigger thermal runaway and may even lead to catastrophic battery fire. On the other hand, when the cooling method is single-phase direct cooling with FC-72 as coolant or two-phase immersed cooling by HFE-7000, the cell temperature is effectively limited to a tolerable level under both high C rate discharging and external shorting conditions. In addition, two-phase immersed cooling scheme is found to lead to better temperature uniformity according to the 75 V battery pack simulations.

Anticipating Poor Outcomes: A Prognostic Machine Learning Model of Unplanned Visits to the Emergency Department for Patients Undergoing Treatment for Head and Neck Cancer Using Comprehensive Multi-Factor Electronic Health Records

Unplanned visits to the emergency department (ED) and hospital admissions (HA) are a significant problem in head neck cancer (HNC) with the literature suggesting that 20-50% patients will visit the ED during active cancer treatment or within 90 days of treatment completion. These unplanned visits compromise outcomes for patients with interruptions or discontinuation of care, and contribute large expense to the healthcare system. Previous work has identified static pretreatment clinical factors associated with ED visits and HA. Here, we used our custom warehouse system to evaluate pre- and during-treatment clinical, dosimetric and toxicity factors.The records of 1285 HNC cancer patients treated in our department between 2014 and 2019 were examined to profile details of a comprehensive set of factors including: demographics, location, diagnosis and staging, chemotherapy, radiation therapy, opioid use, provider reported toxicities and laboratory values and changes to create a prognostic model for ED use during or within 90 days of end of radiation therapy. For continuous variables, Youden's Index was used to identify discriminatory thresholds. Odds ratios were calculated to gauge statistical evidence of relevance. Machine learning models were constructed using gradient boosting (GB), random forest (RF), support vector machines (SVM) and naïve bayes (NB) using 10-fold cross validation.Within our cohort, 17% visited the ED during or within 90 days of the end of radiation therapy. Factors associated with ED visits included female sex (OR = 1.2), single marital status (1.6), smoking history recorded in EHR (1.3), lack of significant alcohol history recorded in EHR (1.3), living within 34 miles of hospital (1.8), age < 59 (1.2), stage IV (1.4), chemotherapy (1.3) including cisplatin (1.5), ≥ 3 changes in opioid Rx (3.4) including morphine (2.0) or fentanyl (2.2), any provider graded toxicity ≥ 3 (2.1). RF had the highest accuracy 0.87 ± 0.005, and AUC of 0.75 ± 0.05. Accuracy was significantly (P < 0.01) less for NB than for RF or GB. Highest importance scoring factors in the model were number of opioid prescription records, body mass index values and changes, volume of low risk PTV volume, age, creatinine, changes in platelet and neutrophil counts, albumin-bilirubin score, and distance.We were able to combine a comprehensive set of clinical, dosimetric and toxicity grading factors to identify an actionable clinical profile for anticipating ED. This forms the basis for next steps to develop and test interventions based on the model as reduction in unplanned acute care is a priority for clinical transformation in oncology. The scope of high-ranking inputs underscores the importance of constructing data ware house systems, that enable constructing multi-domain models. Significance of "distance of hospital" in our model and similar limitations for all ED studies underscores the need for integrated EHR across all hospital systems.

Lumped parameter modeling of two dimensional mechanical and acoustical components

Multidomain lumped parameter models used to predict transducer performance are an efficient way of modeling full system responses. To increase the accuracy of such models, transmission lines can be used. Discretizing the line into small spatial segments allows for lumped parameters to be used for modeling partial differential equations. Segmentation in onedimension has long been used in circuit simulators and multidomain analysis codes. This work presents a solution in two dimensions for both mechanical and acoustical physical domains. For example, the analysis of the mechanical motion of membranes and the acoustical behavior of thin cavities can be analyzed. This enables increased accuracy in modeling a transducer as the shape function of the membrane is accounted for. Further, the mechanical and acoustical physics can be coupled for an appropriate Multiphysics interaction. This is a simple demonstration of the possibilities that could be achieved using a spatially discretized multidomain models for more complex transducers.

Preliminary results from a time domain model of a clarinet-like system

A time domain model of a clarinet-like system has been built using the Simscape multidomain modeling system. This model largely replicates physical features that have been included in previous models by others. The air column has constant inside diameter. The tube has thirteen tone holes. The first three holes are spaced to give approximate chromatic tone spacing. The remaining holes have the same spacing as the last chromatic pair, and all the holes are sized to give a constant tone hole lattice cuttoff frequency of 1500 Hz. The reed is modeled as a single degree of freedom oscillator that terminates the “closed” end of the tube. Blowing pressure is provided by a mouth cavity of fixed volume. This model allows calculation transient behavior that becomes steady state if the input pressure remains constant. We present results of startup transients, slow crescendo and decrescendo playing, and oscillation behavior in response to opening and closing the tone holes. The current model does not accurately match the design of any particular instrument, but it does demonstrate an ability to calculate the behavior of some interesting playing behaviors of an instrument.

Parallel-connected battery module modeling based on physical characteristics in multiple domains and heterogeneous characteristic analysis

Cell inconsistencies inevitably occur inside a battery module. Particularly, the inconsistencies in current distribution and heat generation in a parallel-connected battery module may lead to battery degradation and potential safety issues. Consequently, it is imperative to evaluate and reduce cell inconsistencies in a battery module. In this paper, an extended single particle model of a battery cell is constructed using the Pade approximation and the first-order Taylor expansion to simplify the conventional electrochemical mechanism model. On this basis, a multidomain electrochemical mechanism simulation model of a parallel-connected battery module is attained. Then, the influence of cell inconsistencies on the battery module voltage, internal current distribution and heat generation under different aging situation is assessed by a parameter sensitivity analysis method. Moreover, based on the contribution of each battery model parameter to the inconsistency of the parallel-connected battery module, a battery cell sorting method is proposed. Finally, this proposed sorting method for secondary applications of batteries is validated based on 15 aged batteries. Results indicate that the average standard deviation of the cell current in the NCM parallel-connected module can be reduced from 0.209 A to 0.060 A. The proposed approach is helpful to the fault analysis of electric vehicle battery modules, module level grading or the secondary applications of retired batteries.

VNF-based network service consistent reconfiguration in multi-domain federations: A distributed approach

The Network Function Virtualization (NFV) virtualizes the network appliances — such as routers and firewalls — with software running on commercial off-the-shelf servers. In NFV specification, Network Services (NS) are composed of multiple Virtual Network Functions (VNF) enabling elastic and finer lifecycle management operations such as scaling. For adapting resources to the VNF-based network services evolving context, these operations require to execute, on the fly, a reconfiguration plan supervised by a central orchestrator. This is the single-domain reconfiguration, where the orchestrator has global up-to-date information, ensuring the correct execution of the lifecycle management operations. Moreover, in practice, NS can be implemented by composing VNFs in a cross-domain schema called a multi-domain federation. In this case, the reconfiguration is more challenging since there are multiple orchestrators, one by domain, that manage collaboratively shared network services. Sharing network services creates functional and non-functional dependencies among these services that must be considered for ensuring the consistency of the lifecycle operations. The consistent reconfiguration of shared network services in distributed multi-domain federations is called the NFV Dependent Reconfiguration problem. However, despite being identified as an important challenge by the NFV community, no related solution has yet been proposed. In this paper, we focus on the NFV Dependent Reconfiguration problem. We introduce a distributed approach to guarantee consistency during dependent reconfiguration. The approach is composed of a distributed multi-domain model that establishes the interactions among the federation's entities, and a causally-consistent distributed orchestration algorithm based on such a model. We verify the algorithm viability using, as a case study, the scaling of shared VNF-based network services as defined by the current NFV standard architecture. To the best of our knowledge, the proposed distributed orchestration approach is the first that supports the consistent execution of dependent reconfiguration operations for VNF-based network services.

Polarization switching in Hf0.5Zr0.5O2-dielectric stack: The role of dielectric layer thickness

Understanding the role of the dielectric (DE) layer in ferroelectric (FE) Hf0.5Zr0.5O2 (HZO) based devices (e.g., ferroelectric-field-effect-transistors, FE-FETs) is important to enable their application-driven optimizations. To that end, in this work, we systematically investigate the polarization switching mechanisms in FE–DE stacks and analyze their dependence on the dielectric layer thickness (TDE). First, we fabricate a HZO–Al2O3 (FE–DE) stack and experimentally demonstrate a decrease in remanent polarization and an increase in coercive voltage with an increase in TDE. As such dependencies are out of the scope of commonly used single domain polarization switching models, therefore, we argue that the consideration of the multi-domain model is essential for analyzing the polarization switching in HZO. Then, using phase-field simulations of the FE–DE stack, we show that an increase in TDE results in a larger number of reverse domains in the FE layer to suppress the depolarization field, which leads to a decrease in the remanent polarization and an increase in the coercive voltage. Furthermore, our analysis signifies that the polarization switching mechanism in HZO can be modulated from domain-nucleation based to domain-wall motion based by increasing the TDE and that can serve as a potential knob for application-specific optimization of FE-FETs. In addition, we show that the effective polarization–voltage characteristics of the FE layer in the FE–DE stack exhibit a negative slope region that leads to the charge enhancement effects in the FE–DE stack. While such effects are most commonly misinterpreted as either the transient effects or the stabilized single-domain negative capacitance effects, we demonstrate that the appearance of a negative slope in the hysteretic polarization–voltage characteristics is quasi-static in nature and that originates from the multi-domain polarization switching in the FE.

The Added Value of Sensitivity to Nonnoxious Stimuli to Predict an Individual’s Sensitivity to Pain

Simple tools are needed to predict postoperative pain. Questionnaire-based tools such as the Pain Sensitivity Questionnaire (PSQ) are validated for this purpose, but prediction could be improved by incorporating other parameters. To explore the potency of sensitivity to nonpainful stimuli and biometric data to improve prediction of pain. Transversal exploratory study. Single clinical investigation center. Eighty-five healthy volunteers of both genders underwent a multimodal exploration including biometry, questionnaire-based assessment of anxiety, depression, pain catastrophizing, sensitivity to smell, and the PSQ, followed by a psychophysical assessment of unpleasantness thresholds for light and sound, and sensitivity to mechanical, heat, and cold pain. These last 3 parameters were used to calculate a composite pain score. After a multi-step selection, multivariable analyses identified the explanative factors of experimental pain sensitivity, by including biometric, questionnaire-based, and psychophysical nonnociceptive sensitivity parameters, with the aim of having each domain represented. Female gender predicted mechanical pain, a younger age and dark eyes predicted cold pain, and the PSQ predicted heat pain. Sensitivity to unpleasantness of sound predicted mechanical and heat pain, and sensitivity to unpleasantness of light predicted cold pain. Sensitivity to smell was unrelated. The predictors of the composite pain score were the PSQ, the light unpleasantness threshold, and an interaction between gender and eye color, the score being lower in light-eyed men and higher in all women. The final multivariable multi-domain model was more predictive of pain than the PSQ alone (R2 = 0.301 vs 0.122, respectively). Sensitivity to smell was only assessed by a short questionnaire and could lack relevance. Healthy volunteers were unlikely to elicit psychological risk factors such as anxiety, depression, or catastrophizing. These results have not been validated in a clinical setting (e.g., perioperative). The predictive potential of the PSQ can be improved by including information about gender, eye color, and light sensitivity. However, there is still a need for a technique suitable for routine clinical use to assess light sensitivity.

The Effect of Enzymatic Crosslink Degradation on the Mechanics of the Anterior Cruciate Ligament: A Hybrid Multi-Domain Model

The anterior cruciate ligament’s (ACL) mechanics is an important factor governing the ligament’s integrity and, hence, the knee joint’s response. Despite many investigations in this area, the cause and effect of injuries remain unclear or unknown. This may be due to the complexity of the direct link between macro- and micro-scale damage mechanisms. In the first part of this investigation, a three-dimensional coarse-grained model of collagen fibril (type I) was developed using a bottom-up approach to investigate deformation mechanisms under tensile testing. The output of this molecular level was used later to calibrate the parameters of a hierarchical multi-scale fibril-reinforced hyper-elastoplastic model of the ACL. Our model enabled us to determine the mechanical behavior of the ACL as a function of the basic response of the collagen molecules. Modeled elastic response and damage distribution were in good agreement with the reported measurements and computational investigations. Our results suggest that degradation of crosslink content dictates the loss of the stiffness of the fibrils and, hence, damage to the ACL. Therefore, the proposed computational frame is a promising tool that will allow new insights into the biomechanics of the ACL.