One-step Model Research Articles

Economic system forecasting based on temporal fusion transformers: Multi-dimensional evaluation and cross-model comparative analysis

Although helpful in reducing the uncertainty associated with economic activities, economic forecasting often suffers from low accuracy. Recognizing the high compatibility between deep learning and the nonlinear characteristics of socioeconomic systems, in this paper, we introduce state-of-the-art temporal fusion transformers (TFTs) into the field of economic system forecasting and predict the performance of the Chinese macroeconomic system. Based on an extended analysis of gross final product (GFP) and the intertemporal dynamic relationship between demand-side indicators and output indicators, we establish a scientific economic forecasting framework. To summarize the forecasting characteristics of the TFT algorithm, we compare its one-step and three-step modeling effects in forecasting output indicators with a series of representative benchmark models. According to our proposed four-dimensional evaluation system, the forecasts for China’s macroeconomic system provided by the TFT model have obvious advantages in terms of overall stability, forecasting efficiency, reduction of numerical and timing errors, direction accuracy, and turning point accuracy. The forecast results show that China’s economy faces a risk of slowing growth in the post-pandemic period.

On the critical initiation of planar detonation in Noble-Abel and van der Waals gas

The critical decay rate theory (C.A. Eckett, J.J. Quirk, J.E. Shepherd, Journal of Fluid Mechanics 421 (2000) 147–183) on detonation initiation in unsteady flow was extended from ideal to Noble-Abel and van der Waals gas, in which the finite molecular volume effect (or repulsion force) and the inter-molecular attraction effect (or attraction force) were included. Considering the reactive Euler equations, the reaction zone temperature-gradient equation was established independently of the equation of state and wave geometry, and then solved asymptotically by assuming large activation energy, planar wave front and a one-step irreversible reaction model. A critical decay time was derived as the critical condition for detonation initiation. It was found that the inter-molecular attraction effect makes initiation more difficult by increasing the critical decay time, whereas the finite molecular volume effect promotes initiation. The real gas effects are enhanced when increasing the reduced activation energy and/or decreasing the heat capacity ratio of perfect gas. The applicability and generality of the asymptotic solutions were evaluated by comparison with quasi-unsteady simulation employing detailed reaction models for hydrogen-air and ethylene-air mixtures, and with the further simplified, fully analytical solutions based on strong-shock assumption. The critical conditions of the asymptotic solutions are qualitatively consistent with those obtained with the detailed mechanisms. The asymptotic solutions can also be well reproduced by the simple analytical solutions in most conditions. This study thus establishes a simple, yet reliable, method to evaluate the real gas effect on detonation initiation in high-pressure conditions.

Performance Optimization of Hybrid Draft Biomass Cookstove Using CFD

ABSTRACT The term “hybrid draft biomass cookstove” (HDBC) refers to a stove that employs both natural and forced drafts for its primary and secondary air supply. The HDBC optimization analysis is carried out in this work using computational fluid dynamics (CFD), considering the effect of forced draft secondary air mass flow rates to achieve the optimum (lower emissions and higher efficiency) HDBC performance. The analysis includes combustion and soot modeling to evaluate the thermal and emission performance of the stove. The CFD results are validated against the experimental data obtained from the water boiling test (WBT 4.2.3). The extremely low or high secondary air mass flow rate degrades the stove’s performance; hence, its optimization is essential. The optimal mass flow rate identified in this investigation was 0.017 kg/s, at which the thermal and emission performances of HDBC were evaluated. An eddy dissipation model is employed to simulate biomass combustion, considering the reaction kinetics of wood volatiles reacting with oxygen, and a one-step soot model is used to predict soot formation. The combustion model agreed well with the experimental data for efficiency and temperature, with an average deviance of 10% and 20%, respectively. However, it significantly overestimated CO emissions by 81% and achieved Tier 4 performance in simulation as opposed to Tier 0 in the experiment. The PM2.5 emissions were under predicted by the soot model, but both the CFD and the experiment results achieved Tier 4 performance as per the World Health Organization (WHO) standards. Hence, a one-step soot model was found to be effective in predicting soot formation.

A DNS study of pulverized coal combustion in a hot turbulent environment: Effects of particle size, mass loading and preferential concentration

In the present study, pulverized coal combustion in a hot turbulent environment is investigated using direct numerical simulation (DNS). The coal particles are tracked in the Lagrangian framework while the flow is solved in an Eulerian way. A model involving two competing reactions is used for describing the process of coal devolatilization. The volatile matter is treated as a postulated substance (CaHbOc) and a two-step global reaction mechanism for volatile combustion is adopted. The char combustion is modeled using a one-step global reaction model. The effects of various factors, such as particle size, mass loading and preferential concentration, on the combustion process of coal particles were explored. It was found that, the location of reaction zones appears more upstream and the volatile matter concentration is higher for the cases with smaller particles. The combustion modes were examined. It was shown that the contribution of premixed combustion decreases with increasing particle size while non-premixed combustion prevails in all cases. The increase of particle mass loading results in an increase of volatile matter mass fraction, which leads to higher heat release rate and combustion temperature. Through Voronoï analysis, the preferential concentration of particles was examined. It was found that the preferential concentration behavior of particles is more prominent when the particle Stokes number (St) is close to unity. The volatile matter concentration is higher when particles are preferentially concentrated, facilitating coal particles to ignite.

Circular dichroism in hard X-ray photoelectron diffraction observed by time-of-flight momentum microscopy

X-ray photoelectron diffraction (XPD) is a powerful technique that yields detailed structural information of solids and thin films that complements electronic structure measurements. Among the strongholds of XPD we can identify dopant sites, track structural phase transitions, and perform holographic reconstruction. High-resolution imaging of kll-distributions (momentum microscopy) presents a new approach to core-level photoemission. It yields full-field kx-ky XPD patterns with unprecedented acquisition speed and richness in details. Here, we show that beyond the pure diffraction information, XPD patterns exhibit pronounced circular dichroism in the angular distribution (CDAD) with asymmetries up to 80%, alongside with rapid variations on a small kll-scale (0.1 Å−1). Measurements with circularly-polarized hard X-rays (hν = 6 keV) for a number of core levels, including Si, Ge, Mo and W, prove that core-level CDAD is a general phenomenon that is independent of atomic number. The fine structure in CDAD is more pronounced compared to the corresponding intensity patterns. Additionally, they obey the same symmetry rules as found for atomic and molecular species, and valence bands. The CD is antisymmetric with respect to the mirror planes of the crystal, whose signatures are sharp zero lines. Calculations using both the Bloch-wave approach and one-step photoemission reveal the origin of the fine structure that represents the signature of Kikuchi diffraction. To disentangle the roles of photoexcitation and diffraction, XPD has been implemented into the Munich SPRKKR package to unify the one-step model of photoemission and multiple scattering theory.

Fast automated quantitative phase reconstruction in digital holography with unsupervised deep learning

Digital holography can provide quantitative phase images related to the morphology and content of biological samples. To reconstruct an accurate phase image, several processes, such as phase unwrapping, focusing, and calculation of digital reference wave and numerical propagation, are essential. However, this process is time-consuming. We propose a model that performs phase reconstruction in one-step and two-step using an unsupervised image-to-image translation structure. The two-step reconstruction model translates the phase image, which is obtained by performing numerical propagation on the hologram, into an accurate phase image, whereas the one-step reconstruction model directly translates the hologram into an accurate phase image. The proposed model shows similar high-performance reconstruction to the supervised learning model used in many previous studies. However, since supervised learning is trained in strict pairs, many target domain data (accurate phase imagery) is required. Since the proposed model is trained by unsupervised learning, phase reconstruction can be performed with a small amount of target domain data. The proposed method can help to observe the morphology and movement of biological cells in real-time applications.

Ammonia thermal decomposition on quartz and stainless steel walls

Ammonia has been proposed as a promising solution for hydrogen carriers as well as clean fuels. However, the reaction kinetics is still not robust in many engineering applications, where the deviation is mainly caused by the surface reactions of ammonia on the wall. To examine the ammonia surface reaction on engineering materials, in the present study, the thermal decomposition of ammonia is systematically investigated in uniformly heated quartz/SUS304 tubular flow reactors with the prescribed heating length. A decrease in the inner diameter of the quartz tubular reactor leads to an increase in the ammonia thermal decomposition rate, showing that the thermal decomposition reaction occurs even on quartz surfaces, which is usually believed to be inert. The thermal decomposition on the SUS304 reactor surface starts from as low as 700 K, indicating it is much more reactive than the quartz surface. One-step surface reaction models of ammonia for quartz/SUS304 surfaces are proposed, with which the reaction rates are estimated based on the experimental data. The inner surface of the SUS304 tubular reactor after the thermal decomposition experiments is examined by X-ray photoelectron spectroscopy. The formation of iron nitride on SUS304, which in turn facilitates ammonia thermal decomposition, reveals the positive feedback between ammonia decomposition and nitriding. Moreover, the present one-step surface reaction on stainless steel has been validated by measuring ammonia distributions in the preheat zone of NH3/O2/N2 premixed flame impinging on a stainless steel plate through a two-photon absorption laser-induced fluorescence (TALIF) technique.

Investigation of the wave–photovoltaic-battery​ hybrid power generation platform’s control strategy

This paper proposes a wave–photovoltaic-battery hybrid power generation platform which based on the distributed DC collection and AC inverter grid structure. To be able to simplify the control structure of the power generation system and improve the anti-interference ability of the system, a one-step delay compensation model predictive current control (MPCC) method is proposed for the wave power generation rectifier unit and the perturbation and observation (P&O) fuzzy logic maximum power point tracking (MPPT) algorithm is applied in the photovoltaic power generation unit. Meanwhile, the constant power control method is used to realize the charge and discharge control of the energy storage system to achieve the grid-connected goal of tracking the load power demand. Finally, the MATLAB/Simulink software is applied to simulate the grid-connected operation characteristics of the wave energy power generation unit in continuous power generation modes, as well as the photovoltaic power generation unit at constant temperature and variable irradiance. The results illustrate that the suggested technique for controlling the power generating unit is straightforward and feasible, DC bus voltage stability effect is effective, and the inverter output satisfies the grid-connected operating requirement, which verifies the feasibility of the proposed scheme and can provide some theoretical support for the subsequent high-capacity wave–photovoltaic-battery hybrid power generation platform.

Direct Multi-Material Reconstruction via Iterative Proximal Adaptive Descent for Spectral CT Imaging.

Spectral computed tomography (spectral CT) is a promising medical imaging technology because of its ability to provide information on material characterization and quantification. However, with an increasing number of basis materials, the nonlinearity of measurements causes difficulty in decomposition. In addition, noise amplification and beam hardening further reduce image quality. Thus, improving the accuracy of material decomposition while suppressing noise is pivotal for spectral CT imaging. This paper proposes a one-step multi-material reconstruction model as well as an iterative proximal adaptive decent method. In this approach, a proximal step and a descent step with adaptive step size are designed under the forward-backward splitting framework. The convergence analysis of the algorithm is further discussed according to the convexity of the optimization objective function. For simulation experiments with different noise levels, the peak signal-to-noise ratio (PSNR) obtained by the proposed method increases approximately 23 dB, 14 dB, and 4 dB compared to those of other algorithms. Magnified areas of thorax data further demonstrated that the proposed method has a better ability to preserve details in tissues, bones, and lungs. Numerical experiments verify that the proposed method efficiently reconstructed the material maps, and reduced noise and beam hardening artifacts compared with the state-of-the-art methods.

Geometry-Induced Spin Filtering in Photoemission Maps from WTe_{2} Surface States.

We demonstrate that an important quantum material WTe_{2} exhibits a new type of geometry-induced spin filtering effect in photoemission, stemming from low symmetry that is responsible for its exotic transport properties. Through the laser-driven spin-polarized angle-resolved photoemission Fermi surface mapping, we showcase highly asymmetric spin textures of electrons photoemitted from the surface states of WTe_{2}. Such asymmetries are not present in the initial state spin textures, which are bound by the time-reversal and crystal lattice mirror plane symmetries. The findings are reproduced qualitatively by theoretical modeling within the one-step model photoemission formalism. The effect could be understood within the free-electron final state model as an interference due to emission from different atomic sites. The observed effect is a manifestation of time-reversal symmetry breaking of the initial state in the photoemission process, and as such it cannot be eliminated, but only its magnitude influenced, by special experimental geometries.

A point-wise minimization model for data envelopment analysis considering environmental variables

Environmental variables are widely recognized as a cause of differences in efficiency measurement. However, the existing literature on data envelopment analysis (DEA) in environmental factors ignores the impact of demand on output. To address this gap, we propose the Point-wise Minimization DEA model (PWMDEA), which considers contextual variables that affect demand and lead to differences in efficiency. The model obtains efficiency value by considering the minimum of virtual inputs and virtual demand. Then, efficiency is evaluated by minimizing the ratio of above minimum to virtual output. This one-step model avoids issues of multi-stage assumptions and requires less data, making it more applicable. Moreover, we demonstrate the accuracy of our new model by conducting simulations with given true efficiency values. The simulation results demonstrate that our model has the lowest ranking error when the output is affected by multiple inputs or when demand has a significant impact. In addition, we evaluate the efficiency of healthcare in 31 Chinese provinces by considering two environmental factors. The results suggest that provinces with lower financial investments or population loss received higher rankings from our proposed model. These findings provide plausible explanations and demonstrate the practical usefulness of our model.

CFD simulation of CVD reactors in the CH3SiCl3(MTS)/H2 system using a two-step MTS decomposition and one-step SiC growth models

In this study, we report on a computational fluid dynamics (CFD) simulation of the chemical vapor deposition reactor of silicon carbide (SiC) in the methyltrichlorosilane (MTS, CH3SiCl3)/H2 system. The formation of SiC thin film is controlled by various process parameters, such as temperature and pressure. In this study, we propose a reaction mechanism of MTS decomposition to SiC growth on a substrate surface for CVD reactors in the CH3SiCl3(MTS)/H2 system. The reaction mechanism has two gas-phase pyrolysis reactions and one SiC film formation reaction. However, we individually build and validate MTS decomposition and SiC growth models to reduce uncertainty. An in-house version of reactingFoam, a reactive flow solver within OpenFOAM v2006, was used as the simulation tool. Our model accurately reproduced MTS decomposition for T = 1100–1350 K and [H2]/[MTS] = 2.65–14 at p = 101,325 Pa. Then, the MTS decomposition model was coupled with the SiC growth model, and the coupled model was applied to the SiC deposition data. The model could reproduce multiple datasets through validation studies.

Construction, stochastization and computer study of dynamic population models “two competitors - two migration areas”

When studying deterministic and stochastic population models, the actual problems are the formalization of processes, taking into account new effects caused by the interaction of species, and the development of computer research methods. Computer research methods make it possible to analyze the trajectories of multidimensional population systems. We consider the “two competitors - two migration areas” model, which takes into account intraspecific and interspecific competition in two populations, as well as bidirectional migration of both populations. For this model, we take into account the variability of the reproduction rates of species. A formalized description of the four-dimensional model “two competitors - two migration areas” and its modifications is proposed. Using the implementation of the evolutionary algorithm, a set of parameters is obtained that ensure the coexistence of populations under conditions of competition between two species in the main area, taking into account the migration of these species. Taking into account the obtained set of parameters, a positive stationary state is found. Two-dimensional and three-dimensional projections of phase portraits are constructed. Stochastization of the model “two competitors - two migration areas” is carried out based on the method of self-consistent one-step models constructing. The Fokker-Planck equations are used to describe the structure of the model. A transition to a four-dimensional stochastic differential equation in the Langevin form is performed. To carry out numerical experiments, a specialized software package is used to construct and study stochastic models, and a computer program based on differential evolution is developed. Algorithms for generating trajectories of the Wiener process and multipoint distributions and modifications of the Runge-Kutta method are used. In the deterministic and stochastic cases, the dynamics of the trajectories of populationmigration systems is studied. A comparative analysis of deterministic and stochastic models is carried out. The results can be used in modeling of different classes of dynamic systems.

AiZynthTrain: Robust, Reproducible, and Extensible Pipelines for Training Synthesis Prediction Models.

We introduce the AiZynthTrain Python package for training synthesis models in a robust, reproducible, and extensible way. It contains two pipelines that create a template-based one-step retrosynthesis model and a RingBreaker model that can be straightforwardly integrated in retrosynthesis software. We train such models on the publicly available reaction data set from the U.S. Patent and Trademark Office (USPTO), and these are the first retrosynthesis models created in a completely reproducible end-to-end fashion, starting with the original reaction data source and ending with trained machine-learning models. In particular, we show that employing new heuristics implemented in the pipeline greatly improves the ability of the RingBreaker model for disconnecting ring systems. Furthermore, we demonstrate the robustness of the pipeline by training on a more diverse but proprietary data set. We envisage that this framework will be extended with other synthesis models in the future.

Modeling the growth dependence of Streptococcus thermophilus and Lactobacillus bulgaricus as a function of temperature and pH.

The growth of the lactic acid bacteria (LAB), Streptococcus thermophilus and Lactobacillus bulgaricus, widely used for yogurt production, results in acid production and the reduction of the milk [Formula: see text]. Industrial processes can show temperature ([Formula: see text]) changes due to the large scale of the equipment. As [Formula: see text] and [Formula: see text] affect the LAB growth, this study aimed to model the dependence of S. thermophilus and L. bulgaricus as a function of temperature and pH and to estimate and internally validate their growth parameters and confidence intervals with different modeling approaches. Twenty-four datasets regarding the growth kinetics of S. thermophilus and L. bulgaricus were used for estimating the kinetic parameters for each pure culture. The classical Baranyi and Roberts (sigmoidal) primary and Rosso and coworkers (cardinal parameter) secondary models successfully described the experimental data. The one-step modeling approach showed better statistical results than the two-step approach. The values of eight growth parameters ([Formula: see text], [Formula: see text], [Formula: see text], [Formula: see text], [Formula: see text], [Formula: see text], [Formula: see text], and [Formula: see text]) for each culture estimated from the fitting with the one-step approach and the Monte-Carlo-based approach were similar. Low averaged root-mean-squared errors ([Formula: see text]) (0.125 and 0.090logCFU/mL) and percent discrepancy factor [Formula: see text] ([Formula: see text] and [Formula: see text]) values for S. thermophilus and L. bulgaricus were obtained in the internal model validation, reinforcing the predictive ability of the model.

Numerical and experimental study on the performance of a hybrid draft biomass cookstove

Over 2 billion people worldwide rely on biomass for cooking and primarily use traditional cookstoves, causing 4 million casualties annually due to indoor air pollution. This can be minimised by improving the design and combustion characteristics of the stoves. The present study is an attempt to achieve these improvements by investigating numerically (3D) and experimentally the thermal and emission performance of a newly developed hybrid draft biomass cookstove (HDBC). The influence of secondary air mass flow rate on stove performance is analysed. A homogeneous combustion is modelled to evaluate thermal performance, and soot formation is studied using three soot models: one-step, two-step, and mass-brook. The numerical results are validated against Water Boiling Test (WBT) data. The HDBC achieved Tier 3 level in efficiency and Tier 4 level in CO and PM2.5 emission performance, both experimentally and numerically, thus outperforming standard natural draft stoves. The numerical results exhibited fair agreement with the experimental data, with an average deviation of 6%, 9%, and 26% for efficiency, CO, and PM2.5 emissions, respectively, compared to the experimental results. The soot generation predicted by a one-step soot model gives better results over the other two models and is in good agreement with the experimental results.

Automatic detection of spongiosis associated with oral lichenoid lesions using machine learning

ObjectiveThe purpose of this study was to evaluate whether it is possible to automatically detect spongiosis on whole-slide images (WSIs) derived from patients with either oral lichenoid lesions (OLLs) or oral lichen planus, aimed at reducing pathologist’s diagnostic time and effort. MethodsPatches from WSIs were used to construct a detection model using the VGG16 convolutional neural network. We conducted a one-step method to directly detect spongiosis and a two-step method to detect spongiosis within the epithelium after detection. To compare the impact of imbalanced data, we performed training with and without data augmentation (DA) and random dawn sampling (RD). ResultsFor spongiosis detection, the two-step model with DA and RD obtained an accuracy of 0.97032, a macro-average F1-score of 0.97020, and an area under the receiver operating characteristic curve (AUC) macro-average receiver operating characteristic (ROC) of 0.99513. For the one-step model with DA and RD, the model obtained an accuracy of 0.97388, a macro-average F1-score of 0.97364, and an AUC macro-average ROC of 0.99738. Compared to the oral pathologists marking of WSIs, we noted that the artificial intelligence (AI) model often over diagnosed. ConclusionsThe constructed machine learning classification model for the detection of spongiosis demonstrated good performance, and it is appropriate that oral pathologists use the AI model to suggest results as a reference to reduce the time and effort required for diagnosis.

Sorption isotherm modelling of chilgoza pine nuts (Pinus gerardiana)

Modified forms of the Guggenheim-Anderson-de Boer model (mGAB), Halsey model (mHAL), Henderson model (mHEN) and Oswin model (mOSW) that incorporate a temperature term were selected and used to describe sorption isotherm experimental data of pine nuts that were obtained using the static gravimetric method at 15 °C, 25 °C and 40 °C. A series of goodness-of-fit indices with random and normal residual distributions, the statistical significance and the precision of the parameter estimates were examined to adequately evaluate the pine nut moisture sorption models. The mGAB model best described the sorption isotherms at individual temperatures. Moreover, with the lowest AICc values, the mHEN and mOSW models best differentiated the temperature effect in one-step modelling.

Effects of staggered opposed hot jets on the initiation and propagation of gaseous detonation in a supersonic combustible inflow

In this paper, the detonation initiation mechanism of a supersonic combustible mixture triggered by a staggered opposing combined hot jets was performed. Two-dimensional reactive Navier–Stokes equations with a one-step Arrhenius chemistry model were solved using a structured adaptive mesh refinement framework. The results show that a high temperature and pressure region triggers a rapid detonation initiation after the jet-induced bow shock focusing. Further analysis showed that there is a large baroclinic torque behind the local detonation wave induced by the staggered hot jet, which leads to a large Richtmyer–Meshkov instability at the end of the unburned jet, and the generated periodic shedding vortex structure thereby enhances the diffusion effect in the unburned region. However, the released heat cannot support the propagation of the detonation wave. In addition, different jet intensity distribution schemes and jet spacing will change the ignition point position. It is worth noting that the distance of detonation initiation can be significantly shortened by reducing the front jet intensity while maintaining the total jet energy. Increasing the jet spacing will significantly slow down the detonation initiation process.

Toward a Unified Transformer-Based Framework for Scene Graph Generation and Human-Object Interaction Detection.

Scene graph generation (SGG) and human-object interaction (HOI) detection are two important visual tasks aiming at localising and recognising relationships between objects, and interactions between humans and objects, respectively. Prevailing works treat these tasks as distinct tasks, leading to the development of task-specific models tailored to individual datasets. However, we posit that the presence of visual relationships can furnish crucial contextual and intricate relational cues that significantly augment the inference of human-object interactions. This motivates us to think if there is a natural intrinsic relationship between the two tasks, where scene graphs can serve as a source for inferring human-object interactions. In light of this, we introduce SG2HOI+, a unified one-step model based on the Transformer architecture. Our approach employs two interactive hierarchical Transformers to seamlessly unify the tasks of SGG and HOI detection. Concretely, we initiate a relation Transformer tasked with generating relation triples from a suite of visual features. Subsequently, we employ another transformer-based decoder to predict human-object interactions based on the generated relation triples. A comprehensive series of experiments conducted across established benchmark datasets including Visual Genome, V-COCO, and HICO-DET demonstrates the compelling performance of our SG2HOI+ model in comparison to prevalent one-stage SGG models. Remarkably, our approach achieves competitive performance when compared to state-of-the-art HOI methods. Additionally, we observe that our SG2HOI+ jointly trained on both SGG and HOI tasks in an end-to-end manner yields substantial improvements for both tasks compared to individualized training paradigms.