基于粒计算的多粒度数据分析方法综述

多粒度数据是一种特殊的、有用的数据类型,它通过对论域（研究对象的集合）采用不同的粒化方式使得数据能够在多个粒度空间中进行呈现,在此基础上可以开展数据的多层次知识发现研究。商空间理论、序贯三支决策、多粒度粗糙集、多尺度数据分析模型和多粒度形式概念分析是几种常见的、有效的多粒度数据分析方法,已受到人们的广泛关注。本文对基于粒计算的多粒度数据分析研究工作进行综述,给出每一类多粒度数据分析方法的理论框架、基本概念以及主要研究思想,并指出多粒度数据分析研究中存在的若干问题,为该领域的后续研究提供理论参考。     

Ordered mesoporous Ni/Silica-carbon as an efficient and stable catalyst for CO2 reforming of methane

Ordered mesoporous silica-carbon (MSC) were used as supports of Ni based catalysts for dry reforming of methane (DRM) reaction. The effects of preparation method and precipitant on the catalysts are investigated. The physical and chemical properties are discussed based on the H₂-TPR, FTIR, XRD, TEM, H₂-TPD and N₂ adsorption/desorption characterization. It is found that the preparation method and choice of precipitants affect the catalysts significantly in terms of the properties and catalytic performance in DRM reaction. In detail, the catalysts prepared by the precipitation method show more highly dispersed Ni particles and further better catalytic activity than the impregnated catalyst. That is attributed to the forming Ni₃Si₂O₅(OH)₄ nanoflakes in the catalyst precursors with the existence of alkaline precipitants. And this Ni₃Si₂O₅(OH)₄ species bind the support more tightly than NiO in the impregnated Ni/MSC catalyst. Moreover, the choice of precipitants also influences the form of Ni₃Si₂O₅(OH)₄ species in the catalysts. Specially, the strong electrolytic capacity of NaOH gives the most Ni₃Si₂O₅(OH)₄ nanoflakes formed in Ni-MSC-1 catalyst, which results in the most highly Ni dispersity and further highest catalytic activity. Besides, the strong interaction between the Ni₃Si₂O₅(OH)₄ species and support are also advantageous to the resist sintering and formation of carbon deposition, that is related to the good catalytic stability of catalysts.     

LES study on flow features of the supersonic mixing layer affected by shock waves

The supersonic mixing layer in a model scramjet has been simulated using large eddy simulation (LES) with the consideration of complex shock waves. A comparison with experimental data demonstrates the accuracy and applicability of the employed numerical models and computational methods based on the OpenFOAM solver. The flow features are examined not only from the visualization of flow field, but also from a fundamental point of fluid motion with a focus on evaluating the impacts of shock waves on the turbulent mixing layer. Four basic types of interaction between the shock waves and the mixing layer are discussed according to the flow features and the field development. The mixing properties are analyzed from the evolution of the mixing layer thickness with emphasis on mixing efficiency and total pressure recovery. Results show that three typical developing regions can be observed for the supersonic mixing layer in the scramjet. Shock waves make contributions to the fuel-air mixing due to the amplification of turbulence and the gain of vorticity. The increase in total pressure losses is unavoidable despite the increased mixing efficiency in the presence of shock waves. Additionally, a significant decrease of the convective Mach number is caused by the expansion-fan/shock-wave pattern at the injector exit, leading to a reduction in compressibility effects, which is an important aspect of the physics for the development of the mixing layer in the model scramjet.     

Structural characteristics and microwave dielectric properties of Zn1−xBixVxW1−xO4-based ceramics for LTCC applications

Herein, the improvement of the microwave dielectric properties and sintering characteristics of Zn_1?xBi_xV_xW_1?xO₄(x = 0–0.15)-based ceramics is reported. The results showed that an appropriate amount of doping could not only reduce the optimum sintering temperature from 1100° to 900°C, but also enhance the densification of the microstructures and increase the Q×f value from 5351 to 42525 GHz. Additionally, various structural parameters including the phase composition, crystal structure, vibrational and chemical bond characteristics that are correlated with the dielectric properties were systematically investigated. By considering the chemical bond characteristics, the first-principles calculations and the acquired Raman spectra, the interaction between W-O is stronger than Zn-O in the ZnWO₄ structure, while the interaction between V-O is stronger than Bi-O in BiVO₄. Interestingly, when the Zn_0.97Bi_0.03V_0.03W_0.97O₄-based ceramics were sintered at 900 °C, improved microwave dielectric properties were acquired (ε_r =18.32, Q×f=42525 GHz, τ_f=?67.51 ppm/°C), which provides a promising candidate in low-temperature co-fired ceramics technology.     

Cobalt-doped cerium oxide nanoparticles: Enhanced photocatalytic activity under UV and visible light irradiation

CeO₂ nanoparticles (NPs) were synthesized by coprecipitation using cerium(III) nitrate hexahydrate as the precursor and ethanol as the solvent. Different concentration of cobalt-doped cerium oxide NPs (3mol % and 6 mol %) were prepared by adding various concentrations of cobalt chloride to cerium nitrate. The as-synthesized NPs were characterized through X-ray diffraction (XRD) measurements, ultraviolet (UV)–visible spectroscopy, Photoluminescence (PL) spectroscopy, and transmission electron microscopy (TEM). XRD results reveal that the as-prepared CeO₂ NPs had a face-centered cubic structure with crystallite size in the range of 5–8 nm. TEM analyses showed that the CeO₂ NPs and Co-doped CeO₂ NPs had a homogenous size distribution (sizes were within 5–12 nm). Band-edge absorption of CeO₂ NPs redshifted upon increasing the Co concentration as compared to undoped CeO₂ NPs. PL spectra reveal a peak shift of CeO₂ emission upon cobalt doping, which were due to an increase in oxygen defects localized between the Ce4f and O2p energy levels (i.e., via formation of Ce³⁺ states). Photocatalytic degradation of methylene blue in aqueous solution under UV and visible (sunlight) irradiation in the presence of pure CeO₂ NPs and of Co-doped CeO₂ NPs was investigated. The efficiency of photocatalytic degradation of CeO₂ NPs increased with the Co concentration both under UV irradiation and under visible light. Co-doped CeO₂ NPs (6 mol%) showed degradation efficiencies of 98% and 89% at 420 min of exposure to UV irradiation and to visible light, respectively.     

Proliferative effects of synthetic peptides from β-lactoglobulin and α-lactalbumin on murine splenocytes

Eleven peptides, selected on the basis of physicochemical characteristics and their theoretical release from β-lactoglobulin (β-Lg) and α-lactalbumin (α-La) by trypsin or chymotrypsin, were chemically synthesised to evaluate their immunomodulating properties. Murine splenocyte proliferation in the absence and presence of mitogen and different peptide concentrations were measured after 72–96 h incubations. β-Lg f78–83 had no effect on proliferation; β-Lg f15–20, f55–60, f84–91, f92–105, f139–148, f142–148 and α-La f10–16 stimulated proliferation to different extents; β-Lg f1–8, f102–105 and α-La f104–108 showed a cytotoxic effect. Regression analysis revealed the relationship of positive charge, hydrophobicity and length to the stimulatory proliferative effect. β-Lg f15–20, f55–60 and f139–148 also induced various inhibiting and/or stimulating effects on cytokine secretion. The results confirm that peptides releasable by digestive enzymes from α-La and β-Lg have the potential to influence the specific immune response through the modulation of splenocyte proliferation and cytokine secretion.     

不同解冻方式对生食鱼片解冻品质的影响

为探究不同解冻方式对生食鱼片品质的影响,以虹鳟鱼为研究对象,对冷藏室解冻、水浴解冻、微波解冻和常温空气解冻4 种不同解冻方式下生食鱼片的感官、解冻速率、解冻损失率、色泽、菌落总数、挥发性盐基氮（total volatile basic nitrogen,TVB-N）含量、微观组织结构、质构特性、水分迁移和挥发性气味特征等进行测定。结果表明：冷藏室解冻的生食鱼片感官得分虽显著高于其余3 种解冻方式（P＜0.05）,但其解冻速率明显低于其余3 种解冻方式;水浴解冻组的解冻损失率显著低于其余3 种解冻方式（P＜0.05）,有利于生食鱼片水分的保留,且其TVB-N含量及菌落总数仅较微波解冻组分别高1.02 mg/100 g、0.46（lg（CFU/g））,其感官得分仅较冷藏室解组冻低3.19 分;微波解冻组的白度、菌落总数和TVB-N含量结果虽均优于其他3 种解冻方式,能有效抑制解冻过程中色泽的改变、微生物的滋生和缓解生食鱼片中氨及胺类等碱性含氮物质的产生,但微波解冻使生食鱼片受热不均导致其部分熟化的现象明显,其感官和质构指标均较差,不适宜用于解冻生食鱼片;相较于其他3 种解冻方式,常温空气解冻各个解冻品质指标整体上较差。综上,水浴解冻能最大程度地避免其在解冻中的不良变化,更有利于保证生食鱼片的解冻品质。     

Water management fault diagnosis for proton-exchange membrane fuel cells based on deep learning methods

Water management failure is the most common fault in proton-exchange membrane fuel cells (PEMFCs), and it directly affects the durability and stability of fuel cells. This paper proposes a water fault diagnosis method based on 1DCNN-XGB. To promote its commercial applications, the cathode pressure drop, voltage, and current density were used as the characteristic diagnostic variables,which considered variations in the stack load and also facilitated hierarchical fault diagnosis. First, flooding and drying experiments under different current densities were simulated by changing the operating conditions of the stack, and the obtained experimental data were normalized to eliminate feature imbalances. Then, they were reconstructed into a 1D data set as the input of the model. Automatic feature extraction was performed by a 1DCNN, and the extracted feature maps were used as the input of the XGBoost classifier. Finally, the trained model was validated on the test set for fault diagnosis. The experimental results showed that the model accurately and efficiently distinguished normal and different degrees of two typical fault states (flooding and drying) of the stack with an overall accuracy of 98.10%. The comparative experiments revealed that the model was superior to the individual 1DCNN and XGBoost models, showing greater accuracy and better generalization ability.