Search

Experimental investigations have been conducted in hydrogen-oxygen mixtures with equivalence ratio of 1.5 at cryogenic temperature (77 K) and initial pressure (P 0) from 0.2 to 0.5 atm. The pressure sensors and optical fibers are used to measure the overpressure and flame velocity, respectively. The precursor shock wave is formed by the combination of the 1st and 2nd shock waves. When flame catches up with the precursor shock wave, detonation occurs. Strong flame acceleration was observed in all tested conditions, which can be well characterized and predicted by the Zel'dovich number and expansion ratio. The stuttering and galloping modes were observed at 0.5 and 0.3 atm, respectively. With the decrease of the initial pressure to 0.20–0.25 atm, detonation could not occur, only the deflagration mode was observed. The stability of the mixture can be indicated by the parameter χ , and the improvement of stability of mixture will shorten the initial pressure range where the galloping mode occurs. The heat loss effect on detonation limit has subsequently been examined for the present experiments as well as those with equivalence ratio of 2.6 reported in our previous study [Shen, X. et al., Proceedings of the Combustion Institute 2022]. For equivalence ratio of 2.6, the heat loss is found to have dominant effect on the appearance of detonation limit. Its influence can be quantitatively characterized by critical ratio of heat loss to heat release. However, for equivalence ratio of 1.5, the heat loss effect is found to have only relatively small effect on detonation limit. • The combustion of hydrogen-oxygen mixture at cryogenic temperature of 77 K is investigated. • The 1st and 2nd shock waves would merge to form a stronger precursor shock wave. • Strong flame acceleration could be predicted by expansion ratio and Zel'dovich number. • The stuttering mode and the galloping mode of detonation are both observed. • The effect of heat loss on the detonation limit is examined. [ABSTRACT FROM AUTHOR]

The rate coefficients of H + O 2 = O + OH (R1) were reevaluated using the reported shock tube experimental data from Ryu et al. (1995), Yuan et al. (1991), Masten et al. (1991) and Hong et al. (2011) based on our latest kinetic mechanism [Xueliang Yang et al., Journal of Physical Chemistry A, 2021, 125, 10223–10234]. The statistical analysis using weighted average and weighted least-square regression was performed by applying the updates on specific rates and corresponding uncertainties. The chain-termination reaction of H + O 2 (+ M) = HO 2 (+ M) (R2) was reevaluated as well. New rate formulae of R1 and R2 were derived and incorporated back into the mechanism. The validation demonstrates the updated mechanism has good performance against the measurements such as mass burning rate and ignition delay time at broad range of initial conditions. • The reaction rates of H + O 2 were reevaluated based on shock tube experiments. • Uncertainties of reaction rates were assessed for weighted statistical analysis. • The updated mechanism with reevaluated rates was validated against the measurements. [ABSTRACT FROM AUTHOR]

Syngas is a promising alternative energy carrier with low carbon and pollutants emissions, which has huge application potential in internal combustion engines and gas turbines. The combustion characteristics of stoichiometric syngas/air mixtures with varied initial pressure (0.5–2.5 atm) and hydrogen content (10%–90% v/v) were examined through experiments in a rectangular closed channel. The flame images and overpressure dynamics were captured by high-speed schlieren photography and pressure sensors. The flame instabilities and the flame-acoustic wave interaction were explored. The results showed that the flame morphology and dynamics were enhanced with increasing hydrogen content and initial pressure. The Darrius-Landau instability has an impact on flame deformation, which was strengthened with growing initial pressure, while the thermo-diffusive instability could be neglected during combustion. At later stages after tulip inversion, the flame-acoustic wave interaction was relatively outstanding inducing wrinkled flame front and cellular structures on the flame surface. • Resonance made the flame evolution different at varied initial pressures. • The hydrogen content had greater impact on flame deformation than initial pressure. • Acoustic effects were outstanding in later stages of flame propagation. • Flame-acoustic wave interaction caused cellular structure on the flame surface. [ABSTRACT FROM AUTHOR]

• The lower explosion limit of pulverized coal in syngas/air atmosphere is 50 g / m 3. • Near the lower explosive limit, the addition of syngas increases the maximum explosive pressure of the mixture. • The maximum explosion pressure is 0.81 MPa with pulverized coal concentration of 325 g / m 3. • The volatiles in pulverized coal are mainly involved in combustion reaction. • The oxygen-containing functional groups and aliphatic –CH 3 and –CH 2 are involved in the reaction. To study the explosion characteristics of pulverized coal with syngas/air conditions, micron particle dust explosion experiments are carried out in a 20L spherical explosion vessel. This experiment verifies the safety of syngas carrying pulverized coal transportation when the solid–gas ratio is 300 kg/m3, and the pressure sensor captures the overpressure dynamics. The results show that the explosion intensity increases first and then decreases with the increase of coal dust concentrations, and the maximum value of 0.81 MPa occurs at the concentration of 325 g / m 3. The lower explosion limit (LEL) of coal/syngas/air mixtures is 50 g / m 3 , while the upper explosion limit (UEL) is 4000 g / m 3. The experiments are carried out with varied coal dust concentrations. Scanning electronic microscopy, proximate and FTIR analysis of raw coal and explosive products show that the syngas/air atmosphere can reduce the Lower Explosive Limit of pulverized coal. The volatile matter is the main component involved in the reaction during the explosion of pulverized coal, while the absorption peaks of –OH and aliphatic –CH 3 and –CH 2 in the explosion products are significantly weakened compared with the raw coal. [ABSTRACT FROM AUTHOR]

Deep learning has achieved promising performance in hyperspectral image (HSI) classification. Training deep models usually requires labeling massive HSIs, which, however, is prohibitively time-consuming and expensive. To fill in the gap, this article proposes a novel meta-learning method for HSI few-shot classification that conducts HSI classification with a few labeled samples. Specifically, we introduce the Earth mover’s distance (EMD) as a metric. The designed EMD metric learning module aims to calculate the similarity of paired embedding features by decomposing embedding features into a set of local representations. The EMD metric aims to find the optimal matching flows between local representations that have the minimum matching cost. Furthermore, we attempt to learn class prototype representation for each hyperspectral class using the EMD metric. The proposed network effectively learns general knowledge from base HSIs and transfers such knowledge to the classification of novel HSIs. We conduct HSI few-shot classification by training on three base HSIs and classification on three novel HSIs. Extensive experimental results on three novel HSI datasets demonstrate that the proposed model outperforms the existing state-of-the-art HSI methods, including two HSI few-shot methods. [ABSTRACT FROM AUTHOR]

In the field of image set classification, most existing works focus on exploiting effective latent discriminative features. However, it remains a research gap to efficiently handle this problem. In this paper, benefiting from the superiority of hashing in terms of its computational complexity and memory costs, we present a novel Discrete Metric Learning (DML) approach based on the Riemannian manifold for fast image set classification. The proposed DML jointly learns a metric in the induced space and a compact Hamming space, where efficient classification is carried out. Specifically, each image set is modeled as a point on Riemannian manifold after which the proposed DML minimizes the Hamming distance between similar Riemannian pairs and maximizes the Hamming distance between dissimilar ones by introducing a discriminative Mahalanobis-like matrix. To overcome the shortcoming of DML that relies on the vectorization of Riemannian representations, we further develop Bilinear Discrete Metric Learning (BDML) to directly manipulate the original Riemannian representations and explore the natural matrix structure for high-dimensional data. Different from conventional Riemannian metric learning methods, which require complicated Riemannian optimizations (e.g., Riemannian conjugate gradient), both DML and BDML can be efficiently optimized by computing the geodesic mean between the similarity matrix and inverse of the dissimilarity matrix. Extensive experiments conducted on different visual recognition tasks (face recognition, object recognition, and action recognition) demonstrate that the proposed methods achieve competitive performance in terms of accuracy and efficiency. [ABSTRACT FROM AUTHOR]

Hydrogen storage technology plays an important role on the development of hydrogen fuel cell and lithium aluminum hydride is a powerful candidate for solid-state hydrogen storage materials. However, the high stability and slow dehydrogenation kinetics of LiAlH 4 hinder its application. In this paper, the two-step thermal decomposition properties of LiAlH 4 with and without Fe–Fe 2 O 3 catalysts are investigated. According to the master plots, the model of mample power law (Pn) and nucleation and growth (An) are the optimal mechanism functions for the two-step decomposition of LiAlH 4 , respectively. After doping catalysts, the activation energies decrease significantly. The theoretical and optimized kinetic method are consistent with each other on activation energy. And the latter can also yield other kinetic parameters for a more comprehensive kinetic modelling of LiAlH 4 decomposition. Furthermore, Fe–Al 2 O 3 and Fe–Al intermetallics generated during dehydrogenation might significantly improve the hydrogen release properties of LiAlH 4. • Thermal decomposition kinetics of LiAlH 4 is analyzed using nonisothermal method. • The kinetic parameters are obtained by the optimized kinetic method. • Synergistic catalytic effects of Fe and Fe 2 O 3 composites have been investigated. • Fe–Al 2 O 3 and Fe–Al intermetallics enhance the dehydrogenation properties. [ABSTRACT FROM AUTHOR]

Representing image sets as subspaces on Grassmann manifold and leveraging the Riemannian geometry of this space has proven to be highly effective in various visual recognition tasks. However, the computational cost associated with operating on Riemannian manifolds, especially of high-dimensional ones, poses a significant limitation to the applicability of existing techniques. Moreover, the image sets with labels are scarce and valuable in real-world scenarios, where the manual annotation is time-consuming and costly. To tackle these issues, in this paper, we propose an unsupervised Grassmannian dimensionality reduction method for image set classification and clustering, which simultaneously performs local structure learning and dimensionality reduction under the subspace view. Specifically, instead of relying on the pre-defined similarity matrix that obtained from raw data and may contain much noise, we adaptively learn the local structures from the projected low-dimensional Grassmann manifold. The refined structures then provide more discriminative power to learn the accurate manifold-to-manifold mapping. To better preserve the global structure of the data, this mapping is constrained to satisfy the orthogonal condition. In addition, our model can be compatible with F-norm OLPP for Grassmann manifold when the squared F-norm is transformed into projection metric. A re-weighted iterative scheme is carefully developed to optimize the problem. Experimental results on several benchmark datasets for image set classification and clustering tasks clearly demonstrate that our model consistently outperforms baselines, yielding significant improvements. • We propose a novel unsupervised DR method to work on non-linear Grassmann manifold. • We jointly learn the mapping and local structure to alleviate the effect of noise. • We further measure the distance for Grassmann manifold in terms of F-norm. • The experimental results demonstrate the effectiveness of the proposed model. [ABSTRACT FROM AUTHOR]

The propagation of hydrogen/air premixed flame in channels were investigated, giving particular focus on the effect of aspect ratio through both experimental and numerical investigations. The schlieren system and pressure sensor were used to capture flame shapes and pressure changes, respectively in a channel with aspect ratio of 12. The phenomena of flame shape evolutions, including classic tulip inversion, featured T-shape flame, elongated tulip flame and tulip distortion captured in the experiments, were well reproduced by the numerical simulation. The predicted flame front speed, pressure build-up and their fluctuations are also in reasonably good agreement with the measurements. The validated numerical modelling approach was then applied to investigate the effect of channel aspect ratio on flame shape change, flame propagation and overpressure dynamics. It was found that with increasing aspect ratio, the flame shape changes become more vigorous accompanied by the amplified flame and pressured dynamics due to the enhanced pressure waves, flame-pressure wave interaction and flame instability. The Bychkov's analytical model was used to give predictions on the characteristic times of flame deformation in the channels. By comparison, the analytical model performs better for channel with lower aspect ratio. [Display omitted] • Numerical model was established for propagation in channels. • Aspect ratio effect on hydrogen/air flame evolution in channels was elucidated. • The analytical theory performed better on flame dynamics with lower aspect ratio. [ABSTRACT FROM AUTHOR]

• The OpenFOAM open source software was used to simulate the fine structure of flame propagating in a closed tube. • The effect of pressure field, velocity field, and flow field on hydrogen/air flame was analyzed. • The generation of pressure waves and their interactions with the flame front were analyzed. • The coupling of flow direction, vortex field, and pressure field leads to tulip deformation. • The flame distortion and local flow field caused by the Rayleigh-Taylor instability were analyzed. In this paper, the large eddy simulation (LES) model of OpenFOAM was used to study the flame propagation characteristics of hydrogen-air premixed flame in a closed tube with an aspect ratio of 12. The numerical simulation reproduces the characteristic flame shape evolution observed in the experiment, such as classical tulip flame, T-shaped flame, and distorted tulip deformation. The predicted flame front velocity, overpressure, and its fluctuation are all in good agreement with the measurements. The simulation demonstrates that after the classic tulip flame is formed, a series of waves cause an obvious adverse pressure gradient at the front of the flame. This adverse pressure gradient drives reverse flow which pulls the backward tip further into the combustion zone, forming an elongated tulip flame. A small amount of unburnt gas trapped in the cusp rapidly consumes to form a featured T-shaped flame. Then the T-shaped flame transforms into the second finger-shaped flame, and finally the tulip distortion occurs. The vortex and the reverse flow near the flame front generated by the interaction between the pressure field and the flow field accounts primarily for the formation of distorted tulip flame. [ABSTRACT FROM AUTHOR]

Dust explosions are among the most hazardous incidents in industrial production. The inherent dangers of dust make its testing not only risky but also costly. Consequently, it is crucial to identify a method for rapidly determining dust explosion hazards. Previous quantitative structure-property relationship (QSPR) methods have been computationally demanding, time-consuming, and inefficient. In this study, the minimum ignition energy is predicted using the MPNN model, a graph neural network that extracts molecular data features to make accurate predictions based on chemical structure. The input data is transformed into a compact representation, facilitating more efficient analysis and learning by machine learning algorithms. By converting the molecular structure into SMILES for training, employing molecular enhancement via the SMILES molecular traversal algorithm, and optimizing the model through hyperparameter optimization, the MPNN was compared with the GCN and GAT models. The confusion matrix, ROC curves, and PR curves were evaluated for all three models. The MPNN model outperformed the other two models in predicting dust explosion hazards. After data augmentation and hyperparameter optimization, the performance of all three models improved significantly: the MPNN model enhanced its performance by 4% on the ROC curve and 4% on the PR curve; the GCN model by 6% on the ROC curve and 7% on the PR curve; and the GAT model by 2% on the ROC curve and 3% on the PR curve. Nevertheless, among the optimized models, the MPNN model still demonstrated the best performance on the ROC and PR curves, achieving an accuracy of 98%. By visualizing the classification results using the t-SNE algorithm, it was found that the MPNN model was better at capturing the relationships between nodes on the molecular graph, resulting in superior performance in the classification task. In contrast, the GCN and GAT models underperformed in utilizing graph node information, leading to less efficient predictions. Analysis of the molecular structures in the prediction results indicates that the minimum ignition energy is >30 mJ for chain compounds and below 30 mJ for cyclic compounds, influenced by both the conjugation system and molecular weight. The MPNN model has demonstrated high accuracy and robustness in predicting dust explosion hazards, providing an effective solution to reduce experimental risks and costs. In conclusion, the study presents a method for rapidly determining dust explosion hazards using the MPNN model, a graph neural network that accurately predicts minimum ignition energy based on chemical structure. The MPNN model outperformed other models in terms of accuracy and efficiency, providing an effective solution to reduce experimental risks and costs. These findings have significant implications for industrial production and can be applied at large-scale to enhance safety measures in real-world conditions. [Display omitted] • MPNN model has better performance to predict MIE than GCN and GAT models. • SMILES streamlines geometry optimization and molecular descriptor computation. • Data augmentation and optimization could boost model performance. • Molecular structure has deep correlation with MIE of combustible dusts. • MPNN demonstrates good potential as a complementary tool for MIE investigation. [ABSTRACT FROM AUTHOR]

Sparse representation and dictionary learning have been successfully applied in hyperspectral image classification. Generally, it is more effective to learn the sub-dictionary for each class and utilize multiple scale strategy. However, the sub-dictionary may only consider the within-class information and ignore the discriminative information. For multiscale sparse representation, the shape of the regions would not adaptively change according to context structure. This paper proposes a discriminative sub-dictionary learning algorithm and an adaptive multiscale superpixel classification strategy under sparse representation framework for hyperspectral image classification. In dictionary learning stage, the global constraint term can ensure the learnt sub-dictionaries contain more discriminative information. In sparse representation stage, the adaptive multiscale superpixel strategy can exploit the spatial context according to the structural information. The combination of the two strategies can further explore the spatial and spectral information in hyperspectral image. Experiments on four hyperspectral image datasets prove that the proposed method has a significant improvement in both quantitative and qualitative analysis comparing with state-of-the-art algorithms. Moreover, the proposed algorithm can produce satisfactory results for the imbalanced data problem involved in hyperspectral image. [ABSTRACT FROM AUTHOR]

Riemannian sparse coding methods are attracting increasing interest in many computer vision applications, relying on its non-Euclidean structure. One such recently successful task is image set classification by the aid of Grassmann Manifolds, where an image set can be seen as a point. However, due to irrelevant information and outliers, the probe set may be represented by misleading sets with large sparse coefficients. Meanwhile, it is difficult for a single subspace to cover changes within an image set and the hidden structure among samples is relaxed. In this paper, we propose a novel Grassmann Locality-Aware Group Sparse Coding model (GLGSC) that attempts to preserve locality information and take advantage of the relationship among image sets to capture the inter and intra-set variations simultaneously. Since the contributions of different gallery subspaces to the probe subspace should vary in importance, we then introduce a novel representation adaption term. In addition, a kernelised version of GLGSC is proposed to handle non-linearity in data. To reveal the effectiveness of our algorithm over state-of-the-art, several classification tasks are conducted, including face recognition, object recognition and gesture recognition. [ABSTRACT FROM AUTHOR]

Syngas is a promising alternative energy carrier with low carbon and pollutants emissions, and CO is the main component. The combustion characteristics of stoichiometric CO/air mixtures with varied initial pressure (0.5–1.5 atm) and aspect ratio (12–40) were examined through experiments in a rectangular closed duct. This experiment simulated the transport and storage conditions of CO to study its combustion characteristics at different aspect ratios and initial pressures. The flame images and overpressure dynamics were captured by high-speed camera and pressure sensor, respectively. Experiment results proved that the increase in aspect ratio had an important effect on flame shape evolution. The tilted flame for low aspect ratios was affected by buoyancy, while flames with high aspect ratios were influenced by competing buoyancy and intrinsic instability. The results showed that the flame morphology and dynamics were enhanced with increasing aspect ratio and initial pressure. The intrinsic instability of premixed CO/air combustion was investigated, and the formation mechanism of tilted flame and twisted flame were comprehensively analyzed. Qualitative distinctions were made between symmetric and asymmetric flame morphology regimes. • The cellular flame was only observed at β = 26 and P i > 0.5. • The aspect ratio has more pronounced effect on the flame morphology. • Buoyancy effect and intrinsic instability dominate flame propagation of CO/air mixture. • The morphology regimes of symmetric and asymmetric flames were distinguished. [ABSTRACT FROM AUTHOR]

Experiments were performed in a narrow duct to scrutinize the evolution of premixed stoichiometric hydrogen/air flame propagation at varied initial pressures. Besides the formation of the classic tulip flame, new stages of flame deformation have been observed using high speed Schlieren photography, e.g. elongated tulip flame and T-shape flame. With decreasing initial pressure, the flame propagation tends to be milder, but the classic tulip shape is of robust appearance at all conditions. The measured characteristic times of flame deformation were compared with the predictions of Bychkov's model. The pressure wave generated upon the flame-wall contact at very early stage does not trigger the classic tulip inversion, but will collide with the flame front modifying the tulip structure after reflection on the end wall. Thereafter, three different types of pressure dynamics are initiated depending on the initial pressure, P 0 ≥ 0.7 atm , 0.5 ≤ P 0 < 0.7 atm , or P 0 < 0.5 atm. The Rayleigh–Taylor instability primarily accounts for the flame deformation, whose amplitude growth rate, ω is periodic and determined mainly by the flame acceleration, g according to the linear theory. The measured pressure dynamics by sensor is a combined manifestation originating from the flame dynamics, flame-induced flow and pressure waves excited during combustion. • Premixed hydrogen/air flame propagation in a narrow duct was observed. • New stages of flame deformation were explored. • There are three types of pressure dynamics depending on the initial pressure. • The pressure wave is responsible for the elongation of the classic tulip flame. • The Rayleigh–Taylor instability primarily accounts for the flame deformation. [ABSTRACT FROM AUTHOR]

Hyperspectral imagery (HSI) has shown promising results in real-world applications. However, the technological evolution of optical sensors poses two main challenges in HSI classification: 1) the spectral band is usually redundant and noisy and 2) HSI with millions of pixels has become increasingly common in real-world applications. Motivated by the recent success of hybrid huberized support vector machines (HHSVMs), which inherit the benefits of both lasso and ridge regression, this paper first investigates the advantages of HHSVM for HSI applications. Unfortunately, the existing HHSVM solvers suffer from prohibitive computational costs on large-scale data sets. To solve this problem, this paper proposes simple and effective stochastic HHSVM algorithms for HSI classification. In the stochastic settings, we show that with a probability of at least $1-\varrho $ , our algorithms find an $\epsilon $ -accurate solution using $\tilde {O}({1}/{\lambda _{2}\epsilon })$ iterations. Since the convergence rate of our algorithms does not depend on the size of the training set, our algorithms are suitable for handling large-scale problems. We demonstrate the superiority of our algorithms by conducting experiments on large-scale binary and multiclass classification problems, comparing to the state-of-the-art HHSVM solvers. Finally, we apply our algorithms to real HSI classification and achieve promising results. [ABSTRACT FROM AUTHOR]

Hashing techniques have recently gained increasing research interest in multimedia studies. Most existing hashing methods only employ single features for hash code learning. Multiview data with each view corresponding to a type of feature generally provides more comprehensive information. How to efficiently integrate multiple views for learning compact hash codes still remains challenging. In this article, we propose a novel unsupervised hashing method, dubbed multiview discrete hashing (MvDH), by effectively exploring multiview data. Specifically, MvDH performs matrix factorization to generate the hash codes as the latent representations shared by multiple views, during which spectral clustering is performed simultaneously. The joint learning of hash codes and cluster labels enables that MvDH can generate more discriminative hash codes, which are optimal for classification. An efficient alternating algorithm is developed to solve the proposed optimization problem with guaranteed convergence and low computational complexity. The binary codes are optimized via the discrete cyclic coordinate descent (DCC) method to reduce the quantization errors. Extensive experimental results on three large-scale benchmark datasets demonstrate the superiority of the proposed method over several state-of-the-art methods in terms of both accuracy and scalability. [ABSTRACT FROM AUTHOR]

Multi-view data with each view corresponding to a type of feature generally provides more comprehensive information. Learning from multi-view data is a challenging research topic in pattern recognition. For recognition task, most multi-view learning methods separately learn multi-view dimensionality reduction (MvDR) and classification models. Thus, the connection between the two models has not been well studied. In this paper, we propose a novel multi-view dimensionality reduction and recognition framework, which can establish the connection between MvDR and classification. Specifically, a multi-view dimensionality reduction method, termed as sparse representation regularized multiset canonical correlation analysis (SR 2 MCC) is first proposed. SR 2 MCC considers both correlation and sparse discrimination among multiple views. In accord with SR 2 MCC, a classifier, called multi-view sparse representation based classifier (MvSRC) is further developed. MvSRC performs classification by comparing the reconstruction residuals of different classes among all views. An efficient iterative algorithm is proposed to solve the proposed model. Extensive experiments on the AR, CMU PIE, FERET, and FRGC datasets demonstrate that the proposed framework can achieve superior recognition performance than several state-of-the-art methods. [ABSTRACT FROM AUTHOR]

• Discrete network embedding (DNE) is proposed for compact representation. DNE leverages hash code to represent node, and dramatically reduces computational and storage costs. • Deep discrete attributed network embedding (DDANE) is proposed to effectively leverage node attribute and network structure in attributed network. The proposed DDANE trains an improved graph convolutional network autoencoder to encode node attribute and network structure into latent discrete embedding. • Extensive experiments demonstrate the proposed methods exhibit lower storage and computational complexity than state-of-the-art network embedding methods, and achieve satisfactory accuracy. Network embedding has shown promising performance in real-world applications. The network embedding typically lies in a continuous vector space, where storage and computation costs are high, especially in large-scale applications. This paper proposes more compact representation to fulfill the gap. The proposed discrete network embedding (DNE) leverages hash code to represent node in Hamming space. The Hamming similarity between hash codes approximates the ground-truth similarity. The embedding and classifier are jointly learned to improve compactness and discrimination. The proposed multi-class classifier is further constrained to be discrete to expedite classification. In addition, this paper further extends DNE and proposes deep discrete attributed network embedding (DDANE) to learn compact deep embedding from more informative attributed network. From the perspective of generalized signal smoothing, the proposed DDANE trains an improved graph convolutional network autoencoder to effectively leverage node attribute and network structure. Extensive experiments on node classification demonstrate the proposed methods exhibit lower storage and computational complexity than state-of-the-art network embedding methods, and achieve satisfactory accuracy. [ABSTRACT FROM AUTHOR]

Previous studies have demonstrated that matrix factorization techniques, such as Nonnegative Matrix Factorization (NMF) and Concept Factorization (CF), have yielded impressive results in image processing and data representation. However, conventional CF and its variants with single layer factorization fail to capture the intrinsic structure of data. In this paper, we propose a novel sequential factorization method, namely Graph regularized Multilayer Concept Factorization (GMCF) for clustering. GMCF is a multi-stage procedure, which decomposes the observation matrix iteratively in a number of layers. In addition, GMCF further incorporates graph Laplacian regularization in each layer to efficiently preserve the manifold structure of data. An efficient iterative updating scheme is developed for optimizing GMCF. The convergence of this algorithm is strictly proved; the computational complexity is detailedly analyzed. Extensive experiments demonstrate that GMCF owns the superiorities in terms of data representation and clustering performance. [ABSTRACT FROM AUTHOR]

Hashing techniques have been widely applied in the large-scale cross-view retrieval tasks due to the significant advantage of hash codes in computation and storage efficiency. Most existing cross-view hashing methods can only handle fully-paired scenarios, where all samples from different views are paired. However, such full pairwise correspondences may not be available in practical applications. In this paper, we propose a novel hashing method, named semi-paired hashing (SPH), to deal with a more challenging cross-view retrieval task, where only partial pairwise correspondences are provided in advance. Specifically, SPH aims to preserve within-view similarity and cross-view correlation among multi-view data. Similarity structure within each view is obtained via anchor graph. As limited samples are paired, correlation between unpaired samples is exploited via a simple yet effective approach, which estimates cross-view correlation by partial cross-view pairwise information and within-view similarity structure. Besides, we further incorporate two regression terms between original features and target binary codes to reduce the quantization loss. An efficient iterative algorithm is presented to simultaneously solve hash functions and binary codes. Extensive experiments on two benchmark datasets demonstrate the superiority of SPH over the state-of-the-art methods, especially in the semi-paired scenarios. [ABSTRACT FROM AUTHOR]

The explosion and combustion parameters of mixtures of methane and air with saturated water vapor were experimentally measured in a standard 20-L spherical explosion vessel at room temperature and atmospheric pressure. The experimental data were carefully scrutinized and compared to elucidate the effect of saturated water vapor in air. When the methane content is below 10%, the maximum explosion pressure, maximum rate of pressure rise and laminar flame speed during explosion are slightly affected, which is attributed to the dual role of water vapor: on one hand, the water vapor suppresses the explosion by thermal and chemical effects; on the other hand, the addition of water vapor makes the equivalence ratio higher than that at dry condition with the same methane content. Then when the methane content is relatively higher, the discrepancies between parameters are quite pronounced for the same reason. The flammability limits of methane in the mixture become narrower (6–14%) compared to dry methane/air mixtures (5–15%). Additionally, the measured laminar flame speeds and computed ones by HP-Mech agree well with each other in this work. [ABSTRACT FROM AUTHOR]

In this study, the near-limit behavior of gaseous detonations in three methane–oxygen mixtures (CH 4 –2O 2 , CH 4 –1.5O 2 and CH 4 –4O 2 ) is investigated experimentally. A 36-mm diameter circular tube and three annular channel gaps ( w = 2 mm, 4.5 mm and 7 mm) were used to look at the effect of different geometries on the detonation limits phenomenon. Photodiodes and smoked foils were employed to measure the time-of-arrival of the detonation wave and record the cellular detonation structure, respectively. As the detonation propagates within the limits, the velocity is steady with only a few percent deficit. By decreasing the initial pressure and, hence, reducing the sensitivity of the mixture, the detonation velocity deficit increases gradually. When the initial pressure is approaching the detonation limits, no steady detonation velocity can be realized, and failure occurs. With decreasing initial pressure, the smoked foil records indicate clearly that the cellular detonation evolves from a multi-headed to double-headed and eventually to a single-head spin as the limits approaches. The single-headed spinning detonations for CH 4 –2O 2 , CH 4 –1.5O 2 and CH 4 –4O 2 mixtures in the 36-mm round tube occurs from p 0 = 3–7 kPa, 4–10 kPa and 7–16 kPa, respectively. Using different annular channel widths, it is observed that the detonation velocity decreases as the channel gap is reduced. [ABSTRACT FROM AUTHOR]

Syngas is a promising renewable energy carrier, and its combustion can result in a significant reduction of pollutant emissions. However, syngas has various production sources, its compositions are diverse accordingly. That brings some challenges in its practical application. Therefore, in order to deeply investigate the premixed stoichiometric syngas/air combustion in confined spaces, experiments were conducted in a closed duct with varied aspect ratios (12, 26 and 40) and initial pressures from 0.5 atm to 1.5 atm. Flame shape evolutions in varied cases were captured through high-speed photography, and four flame speed oscillation regimes were distinguished qualitatively. The pressure buildup was recorded by sensors and scrutinized by dimensionless method. The results denoted the propagation dynamics became more violent with higher initial pressure or aspect ratio. Further, the combustion instability was examined, and it indicated the acoustic effect and the Rayleigh-Taylor/Richtmyer-Meshkov instability would be stronger in cases with large aspect ratio, resulting in the deformation of elongated tulip flame. Besides, the Bychkov's theory considering gas compression effect was validated by the experimental data. The discrepancy between them was interpreted with a comprehensive view of enlarged laminar burning speed, non-adiabatic sidewall and finite flame front thickness. [Display omitted] • The aspect ratios (12, 26 and 40) were examined for syngas combustion in ducts. • Flame regimes varied with aspect ratio and initial pressure. • The RT and RM instability has combined effect at later stages. • Bychkov's theory gave better predictions considering gas compression effect. [ABSTRACT FROM AUTHOR]

The premixed hydrogen–air and propane–air flames in a closed duct were experimentally studied for a deep insight into the tulip distortion phenomenon. High speed schlieren photography and pressure sensors recorded the flame images and pressure dynamics at different equivalence ratios for detailed analyses and comparisons. The distorted tulip shape which was first scrutinized and distinguished as an exclusive feature in premixed hydrogen–air flame, was also observed in stoichiometric premixed propane–air flame with a similar behavior on flame shape changes. Tulip distortions are always accompanied by remarkable flame tip velocity fluctuations with near-constant amplitudes. But the pressure dynamics is totally different. There are no observable oscillation and stepped rise in pressure trajectory of stoichiometric premixed propane–air flame. The pressure wave effect is definitely not the incitation factor, but an additional enhancement of flame acceleration, deceleration and deformation. [ABSTRACT FROM AUTHOR]

In this study, a series of experiments have been carried out to determine the detonation velocity deficits of 2H 2 –O 2 and 2H 2 –O 2 –3Ar mixtures in narrow gaps. Those gaps include a 36 mm inner round tube and three annular channels gaps ( w = 2 mm, 4.5 mm and 7 mm, w is the gap of annular channel). Photodiodes and shock pins are employed to measure the time of arrival of the detonation wave. Smoked foils are used to record the cellular detonation structure. Experimental results indicate that for those detonations in 2H 2 –O 2 and 2H 2 –O 2 –3Ar mixtures within the limits, the detonation propagates at a relatively steady velocity with small deficit, as the limits are approached by progressively reducing the initial pressure, the velocity gradually decreases and the failure of detonation occurs as the initial pressure below a critical value. In the 36 mm diameter round tube, the critical pressures for successful propagation detonation in 2H 2 –O 2 and 2H 2 –O 2 –3Ar mixtures are 13 kPa and 11 kPa respectively, and the maximum velocity deficits for above mixtures are 9% and 5%. By introducing a more accurate method that calculate the intermediate parameters (i.e., detonation cell size, reaction zone thickness) for Fay's model, a modified theoretical model to predict the detonation velocity deficits in different geometries is proposed. By comparing with the experimental results, it confirms the theoretical model is suitable for experimental data with reasonable accuracy in most cases. Due to the theoretical model does not consider the boundary condition, which is fairly prominent in very narrow gap (e.g., 2 mm annular channel), therefore, larger discrepancy between the experiment and theoretical prediction can be observed in the case of small gap. Both experimental and theoretical results indicate velocity deficit increases as detonation transmits from round tube to annular channel, which is due to the heat and momentum losses from wall reduces the propagation velocity of detonation. [ABSTRACT FROM AUTHOR]

In this study, an investigation of the explosion parameters, which include maximum explosion pressure ( p max ), maximum rate of pressure rise ( dp / dt ) max , and combustion characteristics, i.e., laminar burning velocity ( S L ), activation temperature ( T a ) and Zel’dovich number ( Ze ) of dimethyl ether (DME)–air–argon/nitrogen mixtures are systematically carried out. The dilution effects on the explosion and combustion characteristics are also explored through the addition of two diluents, i.e., argon (Ar) and nitrogen (N 2 ), into the DME–air mixture. It is found from experiment that, the explosion parameters of fuel-lean and fuel-rich mixture exhibit different behaviors. For the fuel rich mixture, p max decreases with increasing amount of Ar and N 2 dilution, and a significant drop of ( dp / dt ) max is observed as the amount of added diluent increases. For the fuel lean mixture, p max is first increased with increasing diluent, afterwards it decreases as inert gas dilution ratio larger than a critical value, it is also found ( dp / dt ) max fluctuates with more addition of diluent. The present results also confirm that S L decreases more abruptly in the DME–air mixtures with dilution of N 2 than with Ar, which indicates that the increase in density dominates over the retarding effect for laminar burning velocity in the cases of inert gases dilution. The activation temperature of DME–air–Ar is larger than DME–air–N 2 at the same condition, which demonstrates N 2 dilution is more sensitive than N 2 on the suppression of chemical reaction. Results illustrate that Ze goes up with the increase of dilution ratio, and the values of Ze for argon diluted mixtures are larger than that of nitrogen diluted. [ABSTRACT FROM AUTHOR]

Multiset canonical correlation analysis (MCCA) can simultaneously reduce the dimensionality of multimodal data. Thus, MCCA is very much suitable and powerful for multiple feature extraction. However, most existing MCCA-related methods are unsupervised algorithms, which are not very effective for pattern classification tasks. In order to improve discriminative power for handling multimodal data, we, in this paper, propose a unified multiset canonical correlation analysis framework based on graph embedding for dimensionality reduction (GbMCC-DR). Under GbMCC-DR framework, three novel supervised multiple feature extraction methods, i.e., GbMCC-LDA, GbMCC-LDE, and GbMCC-MFA are presented by incorporating several well-known graphs. These three methods consider not only geometric structure of multimodal data but also separability of different classes. Moreover, theoretical analysis further shows that, in some specific circumstances, several existing MCCA-related algorithms can be unified into GbMCC-DR framework. Therefore, this proposed framework has good expansibility and generalization. The experimental results on both synthetic data and several popular real-world datasets demonstrate that three proposed algorithms achieve better recognition performance than existing related algorithms, which is also the evidence for effectiveness of GbMCC-DR framework. [ABSTRACT FROM AUTHOR]

Canonical correlation analysis (CCA) is an efficient method for dimensionality reduction on two-view data. However, as an unsupervised learning method, CCA cannot utilize partly given label information in multi-view semi-supervised scenarios. In this paper, we propose a novel two-view semi-supervised learning method, called semi-supervised canonical correlation analysis based on label propagation (LPbSCCA). LPbSCCA incorporates a new sparse representation based label propagation algorithm to infer label information for unlabeled data. Specifically, it firstly constructs dictionaries consisting of all labeled samples; and then obtains reconstruction coefficients of unlabeled samples using sparse representation technique; at last, by combining given labels of labeled samples, estimates label information for unlabeled ones. After that, it constructs soft label matrices of all samples and probabilistic within-class scatter matrices in each view. Finally, in order to enhance discriminative power of features, it is formulated to maximize the correlations between samples of the same class from cross views, while minimizing within-class variations in the low-dimensional feature space of each view simultaneously. Furthermore, we also extend a general model called LPbSMCCA to handle data from multiple (more than two) views. Extensive experimental results from several well-known datasets demonstrate that the proposed methods can achieve better recognition performances and robustness than existing related methods. [ABSTRACT FROM AUTHOR]

• A novel unsupervised DR approach on Grassmann manifold for image set classification and clustering tasks is proposed. • We propose a Grassmannian unified learning framework to fulfil the gap between similarity learning and projection learning. • An efficient algorithm is proposed to solve the optimization problems which only involve the basic eigenvalue problem. Modeling image sets as points on Grassmann manifold has attracted increasing interests in computer vision community and has been applied to many applications. However, such approaches have suffered from the limitation that high computational cost on Grassmann manifold must be involved, especially high-dimensional ones. In this paper, we propose an unsupervised robust dimensionality reduction algorithm for Grassmann manifold based on Neighborhood Preserving Embedding (GNPE). We first introduce two strategies to construct the coefficients-based similarity graph to eliminate the effects of errors. Then, a projection is learned from the high-dimensional Grassmann manifold to the relative low-dimensional one with more discriminative capability, where the local neighborhood structure is well preserved. To address the issue that the estimated similarity graph is unreliable with noise and outliers, we further propose a unified learning framework which performs similarity learning and projection learning simultaneously. By leveraging the interactions between these two essential tasks, we can capture accurate structures and learn discriminative projections. The proposed method can be optimized by an efficient iterative algorithm. Experiments on various image set classification and clustering tasks clearly show that our model achieves consistent improvements in terms of both effectiveness and efficiency. [ABSTRACT FROM AUTHOR]

A correlation between the hole transport and corresponding structural properties of the bulk regioregular poly(3-hexylthiophene) (rr-P3HT) is studied as a function of temperature by the time-of-flight (TOF) and wide angle X-ray diffraction (WAXD) techniques. The thermally-reversible structural evolution along the (100) and (020) directions in a semi-crystalline rr-P3HT can be divided into two distinct temperature regions. At T > 120 °C, a large thermal expansion along the π-π stacking direction in the nanocrystals and a deteriorated ordering in the material result in negative slopes of temperature and electric field dependences of hole mobility. The WAXD data suggest that the hole transport is limited by a decrease in the crystallinity and by an increase in the hopping distance along the π-π stacking direction, while the Gaussian Disorder Model (GDM) with temperature-independent parameters cannot be applied. At T < 120 °C, the transport-related structural changes are negligible and the temperature and electric field dependences of hole mobility can be described by the GDM with constant energetic (σ ∼ 120 meV) and positional disorder parameters (Σ ∼ 3.33). These values suggest that the hole transport is limited by the amorphous phase, as commonly seen in disordered polymers. Moreover, a regiorandom P3HT (rra-P3HT), which shows a temperature-independent intermolecular distance of ∼15.3Å, provides a route for separate examination of the amorphous phase in rr-P3HT. [ABSTRACT FROM AUTHOR]

Deep insight into each stage of premixed hydrogen-air flame propagation in a horizontal rectangular closed duct was experimentally realized with pressure records and high-speed schlieren photographs under various equivalence ratios. The motion patterns of the flame skirt and the contact point sweeping along the wall were divided into multi-section with distinct features. Two types of tulip cusp motions were scrutinized and distinguished within different equivalence ratio ranges. The vibrations of the flame tip position and velocity were determined as common features in the whole observed range of equivalence ratios. But the tulip distortion is conditional and substantially originates from the vibration just with more remarkable amplitude. The change of pressure difference between the measure points respectively close to the left and right ends of the duct correlates well with the flame tip velocity oscillation with a 90° phase prior. The experimental results were compared with the theoretical prediction. Best agreements are only achieved in a narrow range of equivalence ratio ∅ = 0.91 ∼ 2.24 , which reveals some limitations of the theoretical model by Bychkov et al. [ABSTRACT FROM AUTHOR]

Abstract: Deep insight into each stage of premixed hydrogen-air flame propagation in a horizontal rectangular closed duct was experimentally realized with pressure records and high-speed schlieren photographs under various equivalence ratios. The motion patterns of the flame skirt and the contact point sweeping along the wall were divided into multi-section with distinct features. Two types of tulip cusp motions were scrutinized and distinguished within different equivalence ratio ranges. The vibrations of the flame tip position and velocity were determined as common features in the whole observed range of equivalence ratios. But the tulip distortion is conditional and substantially originates from the vibration just with more remarkable amplitude. The change of pressure difference between the measure points respectively close to the left and right ends of the duct correlates well with the flame tip velocity oscillation with a 90° phase prior. The experimental results were compared with the theoretical prediction. Best agreements are only achieved in a narrow range of equivalence ratio , which reveals some limitations of the theoretical model by Bychkov et al. [Copyright &y& Elsevier]

Exabytes of data are generated daily by humans, leading to the growing needs for new efforts in dealing with the grand challenges for multi-label learning brought by big data. For example, extreme multi-label classification is an active and rapidly growing research area that deals with classification tasks with extremely large number of classes or labels; utilizing massive data with limited supervision to build a multi-label classification model becomes valuable for practical applications, etc. Besides these, there are tremendous efforts on how to harvest the strong learning capability of deep learning to better capture the label dependencies in multi-label learning, which is the key for deep learning to address real-world classification tasks. However, it is noted that there have been a lack of systemic studies that focus explicitly on analyzing the emerging trends and new challenges of multi-label learning in the era of big data. It is imperative to call for a comprehensive survey to fulfil this mission and delineate future research directions and new applications. [ABSTRACT FROM AUTHOR]

Syngas is a promising alternative energy carrier with low pollutant emissions. Carbon monoxide is the main component of syngas. Once accidentally exposed to air, CO/air mixture could be easily ignited starting fire or explosion, which has been scarcely investigated. Therefore, experiments were conducted in a closed duct to study the premixed stoichiometric CO/air combustion characteristics with/without CO 2 addition. Flame images and overpressure were captured by the high-speed schlieren photography and pressure sensor, respectively. It shows that the inhibitor had remarkable negative effect on flame and pressure dynamics. The competitive effect between flame instabilities and buoyancy on flame deformation was established. The classic tulip flame was more robust when the flame instabilities were strengthened along with attenuated buoyancy effect. When the buoyancy effect was significant with higher inhibitor content, an inclined flame front would form without tulip inversion and then propagate downwards. At specific conditions, tulip structure did occur at early stages when the flame instabilities were dominant, while at later stages, with decreasing flame speed, the buoyancy effect became more prominent leading to an inclined front. The thermal and kinetic effects of CO 2 on premixed CO/air combustion were scrutinized for a comprehensive view of inhibiting mechanism. [Display omitted] • CO 2 has significant inhibition on premixed CO/air flame dynamics. • The competition between flame instabilities and buoyancy effect was scrutinized. • The thermal and kinetic effect of CO 2 inhibition on flame was analyzed. [ABSTRACT FROM AUTHOR]

Multi-view subspace clustering aims to learn an appropriate affinity matrix to investigate the relationship between data. However, the learned affinity matrix always has limited representation ability and unclear intrinsic structure. To address the problem, we propose a novel two-step learning strategy to enhance the quality of the affinity matrix. First, we design a generalized framework to calculate the affinity matrix for each view. Unlike most subspace clustering methods that focus on constructing the coefficient matrices, the proposed framework combines the subspace-based and graph-based models to simultaneously learn the coefficient and similarity matrices under a consistency constraint. Thus, the data structure information and multiple views correlation can be captured and propagated between these matrices, which is conducive to generating an informative affinity matrix. Further, we present a post-processing technology called unified good neighbors learning to make the obtained affinity matrix have good sparsity and connectivity. Specifically, we find the good neighbors of each sample by all the views and reassign their coefficients in the final affinity matrix to reflect the essential correlation between samples. Finally, we develop an efficient optimization algorithm and provide the theoretical analyses for the proposed problem. Extensive experiments conducted on six popular multi-view datasets demonstrate the effectiveness of our method. The source code is available at https://github.com/taozhang1223/TAML. • Two-step affinity matrix learning is proposed for multi-view clustering. • The subspace and graph models are combined to capture the data structure. • The unified good neighbors learning is used to reflect the sample correlation. • The efficient algorithms and theoretical analyses are provided for our model. [ABSTRACT FROM AUTHOR]

The Chinese strategic plan "Made in China (2025)" aims at improving the independent innovation capacity of the manufacturing industry. It also commits to constructing green manufacturing systems through cleaner production and green technologies. The Chinese government provides research and development (R&D) preferential policies to promote the realization of these aims. Using the two-stage least square (2SLS) method, this paper tests the impact of R&D capital and its spillover on the energy intensity of the Chinese manufacturing industry. It also evaluates the effectiveness of R&D policy in promoting R&D investment at the industry level. The findings show that R&D capital is positively linked to the reduction of energy intensity, and heterogeneity exists in the relationship between energy intensity and R&D capital across industries. This study finds that in general, R&D subsidies and tax relief do not significantly promote R&D investment, but R&D subsidies generate a significantly positive effect on the R&D intensity in the high-tech industries. • The R&D capital is positively linked to the decline in the energy intensity. • The R&D capital only decreases the energy intensity in energy intensive industries. • In general R&D subsidies and tax relief do not significantly promote R&D intensity. • Public R&D subsidy has a significantly positive effect in high-tech industries. [ABSTRACT FROM AUTHOR]

• The inhibition of N 2 and CO 2 on methane/hydrogen/air premixed flame was investigated. • CO 2 has better inhibiting performance than N 2 at corresponding conditions. • The inhibiting mechanisms were revealed from thermal and kinetic aspects. Hydrogen enriched natural gas (HNG) is a promising alternative fuel. But the blended fuel will inevitably have different ignition and combustion characteristics as compared to natural gas. The extent of the resulting difference depends on the percentage of hydrogen addition. It may affect the compatibility of combustion systems and have safety implications. The present study was aimed at enhancing the safety of HNG through inhibition by inert gases. Laboratory tests were conducted for methane/hydrogen/air premixed flame propagating in a closed channel with either nitrogen (N 2) or carbon dioxide (CO 2) as the inhibitor. Mixtures with different hydrogen volumetric fractions in fuel, including 0%, 20%, 50% or 80% were investigated. The flame shape evolution and the overpressure in the channel were recorded by high-speed Schlieren photography and pressure sensor, respectively. The flame shape was found to change in various ways depending on the inhibitor and hydrogen content. The pressure wave had remarkable impacts on flame and pressure dynamics. The effect of buoyancy on the flame deformation was observed and discussed. Both N 2 and CO 2 were found to have considerable inhibiting effect on the flame speed and overpressure build-up in the channel while the inhibiting effect of CO 2 was stronger. The inhibition mechanisms of either N 2 or CO 2 were revealed from thermal and kinetic aspects. [ABSTRACT FROM AUTHOR]

• Principle component and variation subspaces are constructed for an over-complete dictionary. • A novel prototype and variation model (P+V) based collaborative representation for Grassmann manifolds is proposed to deal with image set classification naturally. • Previous special matrices are generalized to common sparse matrices. • Experimental results show the superiorities of our methods. Image set classification using manifolds is becoming increasingly more attractive since it considers non-Euclidean geometry. However, with the success of dictionary learning for image set classification using manifolds, how to learn an over-complete dictionary is still challenging. This paper proposes a novel prototype subspace learning method, in which a set of images is represented by a linear subspace and then mapped onto a Grassmann manifold. With this subspace representation, class prototypes and intra-class differences can be represented as principal components and variation subspaces, respectively. Isometric mapping further maps the manifolds into the symmetrical space via collaborative representation, which permits a closed-term solution. The proposed method is evaluated for face recognition, object recognition and action recognition. Extensive experimental results on the Honda, Extended YaleB, ETH-80 and Cambridge-Gesture datasets verify the superiority of the proposed method over the state-of-the-art methods. [ABSTRACT FROM AUTHOR]

Background: Periodontal disease, an oral disease characterized by loss of alveolar bone and progressive teeth loss, is the sixth major complication of diabetes. It is spreading worldwide as it is difficult to be cured. The insulin-like growth factor 1 receptor (IGF-1R) plays an important role in regulating functional impairment associated with diabetes. However, it is unclear whether IGF-1R expression in periodontal tissue is related to alveolar bone destruction in diabetic patients. SUMO modification has been reported in various diseases and is associated with an increasing number of biological processes, but previous studies have not focused on diabetic periodontitis. This study aimed to explore the role of IGF-1R in osteogenic differentiation of periodontal ligament stem cells (PDLSCs) in high glucose and control the multiple downstream damage signal factors. Methods: PDLSCs were isolated and cultured after extraction of impacted teeth from healthy donors or subtractive orthodontic extraction in adolescents. PDLSCs were cultured in the osteogenic medium with different glucose concentrations prepared by medical 5% sterile glucose solution. The effects of different glucose concentrations on the osteogenic differentiation ability of PDLSCs were studied at the genetic and cellular levels by staining assay, Western Blot, RT-PCR, Co-IP and cytofluorescence. Results: We found that SNAI2, RUNX2 expression decreased in PDLSCs cultured in high glucose osteogenic medium compared with that in normal glucose osteogenic medium, which were osteogenesis-related marker. In addition, the IGF-1R expression, sumoylation of IGF-1R and osteogenic differentiation in PDLSCs cultured in high glucose osteogenic medium were not consistent with those cultured in normal glucose osteogenic medium. However, osteogenic differentiation of PDLCSs enhanced after adding IGF-1R inhibitors to high glucose osteogenic medium. Conclusion: Our data demonstrated that SUMO1 modification of IGF-1R inhibited osteogenic differentiation of PDLSCs by binding to SNAI2 in high glucose environment, a key factor leading to alveolar bone loss in diabetic patients. Thus we could maximize the control of multiple downstream damage signaling factors and bring new hope for alveolar bone regeneration in diabetic patients. [ABSTRACT FROM AUTHOR]

An experimental study of hydrogen/air premixed flame propagation in a closed rectangular channel with the inhibitions (N 2 or CO 2) was conducted to investigate the inhibiting effect of N 2 and CO 2 on the flame properties during its propagation. Both Schlieren system and the pressure sensor were used to capture the evolution of flame shape and pressure changes in the channel. It was found that both N 2 and CO 2 have considerable inhibiting effect on hydrogen/air premixed flames. Compared with N 2 , CO 2 has more prominent inhibition, which has been interpreted from thermal and kinetic standpoints. In all the flames, the classic tulip shape was observed. With different inhibitor concentration, the flame demonstrated three types of deformation after the classic tulip inversion. A simple theoretical analysis has also been conducted to indicate that the pressure wave generated upon the first flame-wall contact can affect the flame deformation depending on its meeting moment with the flame front. Most importantly, the meeting moment is always behind the start of tulip inversion, which suggests the non-dominant role of pressure wave on this featured phenomenon. • Inhibition effect of N 2 and CO 2 on the hydrogen/air premixed flame was investigated. • Thermal and kinetic inhibiting mechanism of N 2 and CO 2 has been interpreted. • Three types of flame deformation after classic tulip inversion were observed. • Pressure wave plays a non-dominant role on classic tulip inversion. [ABSTRACT FROM AUTHOR]

Interactive image segmentation can extract specific targets meeting users' intention, and has received widespread attention in computer vision. The conventional interactive methods rely too much on the user interaction due to the limitation caused by the hand-crafted low-level features. Recently, deep interactive approaches have significantly improved the segmentation performance thanks to the semantic perception ability. However, in these approaches each interaction is generally treated independently by the same way, regardless of its own intention of each click and the potential relationships among the continuous interactions. The above defects still leads them restricted to the conflict between the interaction quantity and the interaction number. To overcome the above problem, this paper focuses on the click-based interactive segmentation task by explicitly mining the intention of each click and linking the relationships among all clicks. A selection-collection training framework is first established to impose the global object selection and the local error correction roles during the whole interaction process. Then a temporal network architecture is designed to continuously connect the entire click sequence. In this case, the respective role of each click can be played as much as possible, and the spatially varying segmentation cues can be propagated in time series. Experiments on the challenging SBD, GrabCut, DAVIS and Berkeley segmentation datasets demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

Experiments were conducted to investigate the combustion of stacked wet pulverized magnesium. The flame spread over the stack was imaged by infrared camera. The flame shape, flame height, flame spread speed and combustion duration were extracted from infrared images and analyzed. It turned out that the flame spread behavior was closely related to the particle size and moisture content. The smallest particle size at 27 μ m had the most violent combustion compared to 61 μ m and 90 μ m at the same moisture content. With constant particle size, the flame intensity was reinforced as increasing the moisture content reaching the maximum between 20% and 25%. The flame spread would be attenuated instead with higher moisture content. It was believed to be resulted from the competition of chemical (positive) and thermal (negative) effect of water on magnesium combustion. The combustion process and reactions were scrutinized through direct observation and characterization of collected products. The product would either remain original prism-like strip shape or break down to scattered powders. During flame spread, the magnesium would react with oxygen, nitrogen, carbon dioxide and water in air yielding products with disparate colors. Finally, the combustion mechanism was deeply interpreted based on product morphology and reaction analysis. [Display omitted] • The moisture content and particle size had impact on magnesium combustion. • The smallest particle had the most violent combustion at the same moisture content. • There was competition between chemical and thermal effect of water on magnesium combustion. • The products were characterized to interpret the reaction pathways. • The combustion mechanism was proposed based on product morphology and reactions. [ABSTRACT FROM AUTHOR]

This paper reports the results of an experimental study of the velocity fluctuation near the detonation limits for CH 4 2H 2 3O 2 mixture in tubes with a length ( L t ) of 2.5 m and diameter ( d ) of 4-mm, 14-mm and 36-mm. Fiber optics are used to measure the local velocity along the test section. Results suggest that the velocity behavior is complicated in 4-mm diameter tube (with L t / d = 625) at the condition near the detonation limits. As the initial pressure reduced from 55 kPa to as low as 16 kPa, it can be observed six propagation modes with different features, these are steady detonation, fast fluctuation-stable detonation, stuttering detonation, stuttering detonation-fast deflagration, fast deflagration, and flame propagation. In 14-mm diameter tube ( L t / d = 178.6), it can be observed three propagation modes: steady detonation, galloping detonation and flame. Only steady detonation and flame propagation modes exist in the 36-mm diameter tube ( L t / d = 69.4). Results indicate that initial and boundary conditions have strong influences on the propagation of the detonation wave in small-diameter tube. The boundary layer displacement thickness is more significant in the smaller tube and lower initial pressure, which causes more losses through the flow divergence and therefore the increasing of velocity deficits. It is believed that the competition between the losses from the tube wall and the detonation instability causes the complicated velocity fluctuation in the smaller diameter tube as it is approaching the propagation limits. Results also confirm d / λ is an appropriate length scale for determining detonation limit. [ABSTRACT FROM AUTHOR]

In this study, detonation cell sizes of binary fuel blends of methane–hydrogen–oxygen mixtures with different compositions (i.e., CH 4 –2H 2 –3O 2 , CH 4 –H 2 –2.5O 2 and CH 4 –4H 2 –4O 2 ) are experimentally measured. Good agreement is found between the experimental data and predictions based on the chemical length scales obtained from a detailed chemical kinetic model. The results show that detonation structure in methane–hydrogen–oxygen mixtures are irregular and detonation cell size ( λ ) for CH 4 –H 2 –2.5O 2 mixture is larger than CH 4 –2H 2 –3O 2 and CH 4 –4H 2 –4O 2 mixtures. A linear relationship is determined by: λ = 34.62 Δ I , through scaling the detonation cell size with the ZND induction zone length ( Δ I ) for methane–hydrogen–oxygen mixtures, and this scaling function is universal in CH 4 –H 2 –O 2 mixtures at three compositions. [ABSTRACT FROM AUTHOR]