Experimental research on combined effect of obstacle and local spraying water fog on hydrogen/air premixed explosion

In order to reveal the mechanism of water fog explosion suppression and research the combined effect of water fog and obstacle on hydrogen/air deflagration, multiple sets of experiments were set up. The results show that the instability of thermal diffusion under lean combustion conditions is the main influencing factor of hydrogen/air flame surface instability, and the existence of water fog will aggravate the hydrogen/air flame surface instability. When obstacle is not considered, 8 μm, 15 μm, 30 μm water fog can significantly reduce the flame velocity and explosion overpressure of hydrogen/air, 45 μm fine water fog plays the opposite role. When considering the relative position of the water fog release position and the obstacle, the 8 μm, 15 μm, 30 μm water fog has almost no suppression effect when released near the obstacle, but a significant suppression effect occur, when using the 45 μm water fog. In the field of theoretical research, the research results not only provide an experimental basis for the fine water fog to reduce the consequences of hydrogen explosion accidents, and the optimal diameter range used by the water fog, but also provide experimental reference for the numerical simulation of hydrogen/air explosion suppression in semi-open space, and promote the development of hydrogen explosion suppression theory. In terms of engineering applications, this study can provide a theoretical basis for the layout of fire fighting equipment in the engine room of nuclear power plants or hydrogen-powered ships.     

Tabular data: Deep learning is not all you need

A key element in solving real-life data science problems is selecting the types of models to use. Tree ensemble models (such as XGBoost) are usually recommended for classification and regression problems with tabular data. However, several deep learning models for tabular data have recently been proposed, claiming to outperform XGBoost for some use cases. This paper explores whether these deep models should be a recommended option for tabular data by rigorously comparing the new deep models to XGBoost on various datasets. In addition to systematically comparing their performance, we consider the tuning and computation they require. Our study shows that XGBoost outperforms these deep models across the datasets, including the datasets used in the papers that proposed the deep models. We also demonstrate that XGBoost requires much less tuning. On the positive side, we show that an ensemble of deep models and XGBoost performs better on these datasets than XGBoost alone.     

Effects of high-pressure-modified soy 11S globulin on the gel properties and water-holding capacity of pork batter

The effects of high-pressure-modified soy 11S globulin (0.1, 200, and 400 MPa) on the gel properties, water-holding capacity, and water mobility of pork batter were investigated. The high-pressure-modified soy 11S globulin significantly increased (P < 0.05) the emulsion stability, cooking yield, hardness, springiness, chewiness, resilience, cohesiveness, the a^* and b^* values, and the G′ and G′′ values of pork batter at 80 °C, compared with those of 0.1 MPa-modified globulin. In contrast, the centrifugal loss and initial relaxation time of T_2b, T₂₁, and T₂₂ significantly decreased (P < 0.05). Meanwhile, the microstructure was denser, and the voids were smaller and more uniform compared with those of 0.1 MPa-modified globulin. In addition, the sample with 11S globulin modified at 400 MPa had the best water-holding capacity, gel structure, and gel properties among the samples. Overall, the use of high-pressure-modified soy 11S globulin improved the gel properties and water-holding capacity of pork batter, especially under 400 MPa.     

Data fusion and transfer learning empowered granular trust evaluation for Internet of Things

In the Internet of Things (IoT), a huge amount of valuable data is generated by various IoT applications. As the IoT technologies become more complex, the attack methods are more diversified and can cause serious damages. Thus, establishing a secure IoT network based on user trust evaluation to defend against security threats and ensure the reliability of data source of collected data have become urgent issues, in this paper, a Data Fusion and transfer learning empowered granular Trust Evaluation mechanism (DFTE) is proposed to address the above challenges. Specifically, to meet the granularity demands of trust evaluation, time–space empowered fine/coarse grained trust evaluation models are built utilizing deep transfer learning algorithms based on data fusion. Moreover, to prevent privacy leakage and task sabotage, a dynamic reward and punishment mechanism is developed to encourage honest users by dynamically adjusting the scale of reward or punishment and accurately evaluating users’ trusts. The extensive experiments show that: (i) the proposed DFTE achieves high accuracy of trust evaluation under different granular demands through efficient data fusion; (ii) DFTE performs excellently in participation rate and data reliability.     

Numerical simulation of water and heat transport in the cathode channel of a PEM fuel cell

Cathode channel of a PEM fuel cell is the critical domain for the transport of water and heat. In this study, a mathematical model of water and heat transport in the cathode channel is established by considering two-phase flow of water and air as well as the phase change between water and vapor. The transport process of the species of air is governed by the convection-diffusion equation. The VOSET (coupled volume-of-fluid and level set method) method is used to track the interface between air and water, and the phase equilibrium method of water and vapor is employed to calculate the mass transfer rate on the two-phase interface. The present model is validated against the results in the literature, then applied to investigate the characteristics of two-phase flow and heat transfer in the cathode channel. The results indicate that in the inlet section, water droplets experience three evolution stages: the growing stage, the coalescence stage and the generation stage of dispersed water drops. However, in the middle and outlet sections of the channel, there are only two stages: the growth of water droplets, and the formation of a water film. The mass transfer rate of phase change in the inlet section of the channel varies over time, exhibiting an initial increase, a decrease followed, and a stabilization finally, with the maximum and stable values of 1.78 × 10^?4 kg/s and 1.52 × 10^?4 kg/s for Part 1, respectively. In the middle and outlet sections, the mass transfer rate increase firstly and then keeps stable gradually. Furthermore, regarding the distribution of the temperature and vapor mass fraction in the channel, near the upper surface of the channel, the temperature and vapor mass fraction first change slightly (x < 0.03 m) and then rapidly decrease with fluctuations (x > 0.03 m). In the middle of the channel, the temperature and vapor mass fraction slowly decrease with fluctuation.     

Learn decision trees with deep visual primitives

In this paper, we strive to propose a self-interpretable framework, termed PrimitiveTree, that incorporates deep visual primitives condensed from deep features with a conventional decision tree, bridging the gap between deep features extracted from deep neural networks (DNNs) and trees’ transparent decision-making processes. Specifically, we utilize a codebook, which embeds the continuous deep features into a finite discrete space (deep visual primitives) to distill the most common semantic information. The decision tree adopts the spatial location information and the mapped primitives to present the decision-making process of the deep features in a tree hierarchy. Moreover, the trained interpretable PrimitiveTree can inversely explain the constituents of the deep features, highlighting the most critical and semantic-rich image patches attributing to the final predictions of the given DNN. Extensive experiments and visualization results validate the effectiveness and interpretability of our method.     

An infinite chain, {[Ni(tn)2]3[Fe(CN)4(μ-CN)2]2}n,a new catalyst for electrochemical-driven water splitting and photochemical-driven hydrogen evolution from water under blue light

A new catalyst for both water reduction and oxidation, based on an infinite chain, {[Ni(tn)₂]₃ [Fe(CN)₄ (μ-CN)₂]₂}_n, is formed by the reaction of NiCl₂, 1,3-propanediamine (tn) and K₃ [Fe(CN)₆]. {[Ni(tn)₂]₃ [Fe(CN)₄ (μ-CN)₂]₂}_n can electro-catalyze hydrogen evolution from a neutral aqueous buffer (pH 7.0) with a turnover frequency (TOF) of 1561 mol of hydrogen per mole of catalyst per hour (H₂/mol catalyst/h) at an overpotential (OP) of 837 mV {[Ni(tn)₂]₃ [Fe(CN)₄ (μ-CN)₂]₂}_n also can electro-catalyze O₂ production from water with a TOF of ～45 mol O₂ (mol cat)^?1s^?1 at an OP of 591 mV. Under blue light (λ = 469 nm), together with CdS nanorods (CdS NRs) as a photosensitizer, and ascorbic acid (H₂A) as a sacrificial electron donor, {[Ni(tn)₂]₃ [Fe(CN)₄ (μ-CN)₂]₂}_n can photo-catalyze hydrogen generation from an aqueous buffer (pH 4.0) with a turnover number (TON) of 11,450 mol H₂ per mole of catalyst (mol of H₂ (mol of cat)^?1) during 10 h irradiation. The average of apparent quantum yield (AQY) is as high as 40.96% during 10 h irradiation. Studies indicate that {[Ni(tn)₂]₃ [Fe(CN)₄ (μ-CN)₂]₂}_n exists in two forms: a cyano-bridged chain ({[Ni(tn)₂]₃ [Fe(CN)₄ (μ-CN)₂]₂}_n) in solid, and a salt ([Ni(tn)₂]₃ [Fe(CN)₆]₂) in aqueous media; Catalytic reaction occurs on the nickel center of [Ni(tn)₂]²⁺, and the introduction of [Fe(CN)₆]^3- can improve the catalytic efficiency of [Ni(tn)₂]²⁺ for H₂ or O₂ generation. We hope these findings can afford a new method for the design of catalysts for both water reduction and oxidation.     

A data-driven digital-twin prognostics method for proton exchange membrane fuel cell remaining useful life prediction

Prognostics and health management of proton exchange membrane fuel cell (PEMFC) systems have driven increasing research attention in recent years as the durability of PEMFC stack remains as a technical barrier for its large-scale commercialization. To monitor the health state during PEMFC operation, digital twin (DT), as a smart manufacturing technique, is applied in this paper to establish an ensemble remaining useful life prediction system. A data-driven DT is constructed to integrate the physical knowledge of the system and a deep transfer learning model based on stacked denoising autoencoder is used to update the DT with online measurement. A case study with experimental PEMFC degradation data is presented where the proposed data-driven DT prognostics method has applied and reached a high prediction accuracy. Furthermore, the predicted results are proved to be less affected even with limited measurement data.     

Dysprosium doped copper oxide (Cu1-xDyxO) nanoparticles enabled bifunctional electrode for overall water splitting

The production of hydrogen, a favourable alternative to an unsustainable fossil fuel remains as a significant hurdle with the pertaining challenge in the design of proficient, highly productive and sustainable electrocatalyst for both oxygen evolution reaction (OER) and hydrogen evolution reaction (HER). Herein, the dysprosium (Dy) doped copper oxide (Cu_1-xDy_xO) nanoparticles were synthesized via solution combustion technique and utilized as a non-noble metal based bi-functional electrocatalyst for overall water splitting. Due to the improved surface to volume ratio and conductivity, the optimized Cu_1-xDy_xO (x = 0.01, 0.02) electrocatalysts exhibited impressive HER and OER performance respectively in 1 M KOH delivering a current density of 10 mAcm^?2 at a potential of ?0.18 V vs RHE for HER and 1.53 V vs RHE for OER. Moreover, the Dy doped CuO electrocatalyst used as a bi-functional catalyst for overall water splitting achieved a potential of 1.56 V at a current density 10 mAcm^?2 and relatively high current density of 66 mAcm^?2 at a peak potential of 2 V. A long term stability of 24 h was achieved for a cell voltage of 2.2 V at a constant current density of 30 mAcm^?2 with only 10% of the initial current loss. This showcases the accumulative opportunity of dysprosium as a dopant in CuO nanoparticles for fabricating a highly effective and low-cost bi-functional electrocatalyst for overall water splitting.