基于计算机视觉技术的茶叶品质随机森林感官评价方法研究

为弥补茶叶品质感官审评存在的缺陷，利用计算机视觉技术对茶叶品质进行快速无损评价研究。以碧螺春绿茶为对象，依据专家感官审评结果，将茶样分成4个等级；采用中值滤波及拉普拉斯算子对茶样图像进行预处理，并提取预处理后的茶样图像的颜色特征和纹理特征以表征茶叶图像的外形特征，利用随机森林算法对茶叶外形特征属性进行重要性排序；筛选出重要性较大的特征及随机森林算法中最优的决策树棵数建立感官评价模型，并与建立的支持向量机(SVM)模型性能相比较。结果表明：色调均值、色调标准差、绿体均值、平均灰度级、饱和度均值、红体均值、饱和度标准差、亮度均值、一致性等9个特征属性的重要性较大，且与感官审评特征描述结果相一致；当采用优选出的9个重要性较大的特征及决策数棵数为500时，建立的模型性能最优，模型总体判别率为95.75%，Kappa系数为0.933，OOB误差为5%，较SVM模型分别提高了3.5%，0.066，优选的9个重要性较大的图像特征与感官审评特征描述相一致。研究表明：利用随机森林方法筛选出对茶叶外形特征属性贡献最大的少数几个特征建立模型，模型性能就能达到很好的识别效果，模型得到简化，同时模型精度和稳定性都高于其他方法。     

Two weight commutators in the Dirichlet and Neumann Laplacian settings

In this paper we establish the characterization of the weighted BMO via two weight commutators in the settings of the Neumann Laplacian ${\mathrm{\Delta }}_{N_{+}}$ on the upper half space ${\mathbb{R}}_{+}^n$ and the reflection Neumann Laplacian ${\mathrm{\Delta }}_N$ on ${\mathbb{R}}^n$ with respect to the weights associated to ${\mathrm{\Delta }}_{N_{+}}$ and ${\mathrm{\Delta }}_N$ respectively. This in turn yields a weak factorization for the corresponding weighted Hardy spaces, where in particular, the weighted class associated to ${\mathrm{\Delta }}_N$ is strictly larger than the Muckenhoupt weighted class and contains non-doubling weights. In our study, we also make contributions to the classical Muckenhoupt–Wheeden weighted Hardy space (BMO space respectively) by showing that it can be characterized via the area function (Carleson measure respectively) involving the semigroup generated by the Laplacian on ${\mathbb{R}}^n$ and that the duality of these weighted Hardy and BMO spaces holds for Muckenhoupt $A^p$ weights with $p\in (1,2]$ while the previously known related results cover only $p\in (1,\frac{n+1}{n}]$. We also point out that this two weight commutator theorem might not be true in the setting of general operators L, and in particular we show that it is not true when L is the Dirichlet Laplacian ${\mathrm{\Delta }}_{D_{+}}$ on ${\mathbb{R}}_{+}^n$.     

Proton conductivity study on three CS/IL@fle-MOF membranes

The low-cost, high specific surface area and porosity, controlled pore size, and chemical properties of metal–organic framework (MOF) materials have attracted much attention in the exploration of proton conduction. The method of chemically modifying MOF structures or introducing conductive medium into the holes can effectively improve the proton conductivities of the materials. Here, the structural tunability of ionic liquid (IL) and flexible MOF (fle-MOF) materials are matched to give full play to the conductivity of IL, the framework support, and the microporous effect of MOFs, which achieves the synergistic effect of performance and expands the temperature range of proton transfer. Three kinds of CS/IL@fle-MOF membranes were prepared by combining three fle-MOFs with 1-carboxymethyl-3-methylimidazole (CMMIM) in different proportions to obtain 15 pieces of membranes. The comparative analyses show that CS/IL@fle-MOF membranes have excellent proton conduction performance at a wider temperature range (263–353 K) and lower relative humidity (75% RH). Among them, the proton conductivities of CS/CMMIM@MIL-88A-25% and CS/CMMIM@MIL-88B-125% are up to 1.33 and 1.42 S cm⁻¹ at 75% RH and 353 K, respectively; whereas those of CS/CMMIM@MIL-53(Fe)-75% and CS/CMMIM@MIL-88B-125% reach up to 2.1 × 10⁻³ and 1.28 × 10⁻³ S cm⁻¹ at 75% RH and 263 K, respectively. The E_a of CS/CMMIM@fle-MOFs is in the range of 0.1–0.5 eV, suggesting that the proton transport follows predominantly the typical Grotthuss transfer mechanism. The results of this study indicate that the CS/CMMIM@fle-MOF membranes combinations offer great potential for the design of composite porous proton-conducting materials.     

Athermal and Soft Multi-Nanopatterning of Azopolymers: Phototunable Mechanical Properties

Imprinting nanopatterns on flexible substrates has diverse applications in advanced fabrication. However, the traditional thermal nanoimprint lithography (T-NIL) often causes shrinkage upon cooling. Here, a simple yet versatile method is introduced to fabricate multiple nanopatterns on a flexible substrate coated with an azopolymer by combining athermal nanoimprint lithography (AT-NIL) and photolithography. The azopolymer has various mechanical properties upon photoirradiation: 1) phototunable glass-transition temperatures (T_g) and concomitantly photoinduced switch from glassy plastic to viscoplastic polymer; 2) prominent modulation of viscoplasticity under light illumination at different wavelengths. Regionally selective multiple nanopatterns are conveniently fabricated, presenting angle-dependent structural color images on poly(ethylene terephthalate) (PET) substrates. The flexible, athermal and multiple nanopatterning method has the potential for on-demand fabrication of complex nanopatterns.     

Possible light neutral boson and particle mass quantization

Qualities of nucleons, such as the fundamental parameter mass, might be modified in extreme conditions relative to those of isolated nucleons. We show the ratio of the EMC-effect tagged nucleon mass to that of the free one(m*/m);these values are derived from the nuclear structure function ratio between heavy nuclei and deuterium measured in the electron Deep Inelastic Scattering(DIS) reaction in 0.3≤x≤0.7. The increase in m*/mwith A-1/3 is phenomenologically interpreted via the release of a color-singlet cluster formed by sea quarks and gluons in bound nucleons holding high momentum in the nucleus, from which the mass and fraction of non-nucleonic components in nuclei can be deduced. The mass of color-singlet clusters released per short range correlated(SRC) proton in the high momentum region(k > 2 fm-1) is extracted to be 16.890 ±0.016 MeV/c2, which evidences the possibility of a light neutral boson and quantized mass of matter.     

Study on the mechanism of catalytic synthesis of dimethyldichlorosilane by AlCl3/MIL‐53(Al)@γ‐Al2O3

Dimethyldichlorosilane, one of the most consumed organosilicon monomers in the industry, can be prepared in a highly efficient and environmentally friendly synthesis method of disproportionating methylchlorosilanes. However, the internal mechanism of the reaction remains unclear. In this paper, the mechanism catalyzed by AlCl₃/MIL‐53(Al) and AlCl₃/MIL‐53(Al)@γ‐Al₂O₃ catalysts was calculated at B3LYP/6‐311++G(3df, 2pd) level by using the density functional theory (DFT). The results showed that although the two catalysts had similar active structures, the catalytic effects were significantly different. The Lewis acid center on the surface of γ‐Al₂O₃ in the core‐shell catalyst is complementary to the classic Lewis acid AlCl₃ through the spatial superposition effect, which greatly improves the Lewis acid catalytic activity of AlCl₃/MIL‐53(Al)@γ‐Al₂O₃.     

污泥基吸附剂对亚甲基蓝和环丙沙星的吸附性能研究

亚甲基蓝和环丙沙星是水体中2种污染物, 对生态环境有潜在危害. 本文以市政剩余活性污泥为原料, 氯化锌为活化剂热解制备污泥基吸附剂, 研究盐酸酸洗浓度、氯化锌浓度、热解温度、热解时间等对污泥基吸附剂吸附水中亚甲基蓝和环丙沙星性能的影响. 结果表明 (1)污泥基吸附剂对亚甲基蓝的吸附性能随盐酸酸洗浓度的增大而增加, 对环丙沙星的吸附性能则随盐酸酸洗浓度的增大呈先降后增趋势, 两者均在1.500mol·L-1盐酸浓度下取得最优值. (2)污泥基吸附剂对亚甲基蓝和环丙沙星的吸附性能随氯化锌浓度和热解温度的增加呈先升后降趋势, 在氯化锌浓度为4.0mol·L-1、热解温度为500℃时有最优值; 随着热解时间的延长, 污泥基吸附剂对亚甲基蓝和环丙沙星的吸附性能分别在500℃热解70min和80min时有最优值. (3)污泥基吸附剂的最佳制备条件为 氯化锌4.0mol·L-1活化2h、500℃热解70min和80min、1.500mol·L-1盐酸酸洗; 以此制得的污泥基吸附剂对亚甲基蓝和环丙沙星的去除率分别为97.7%和96.4%, 平衡吸附量分别为97.9mg·g-1和3.9mg·g-1, 且污泥基吸附剂对亚甲基蓝和环丙沙星的吸附过程均符合准二级动力学方程.     

Generalized Jacobi spectral Galerkin method for fractional pantograph differential equation

This work is concerned with the extension of the Jacobi spectral Galerkin method to a class of nonlinear fractional pantograph differential equations. First, the fractional differential equation is converted to a nonlinear Volterra integral equation with weakly singular kernel. Second, we analyze the existence and uniqueness of solutions for the obtained integral equation. Then, the Galerkin method is used for solving the equivalent integral equation. The error estimates for the proposed method are also investigated. Finally, illustrative examples are presented to confirm our theoretical analysis.     

Enhanced thermal conductivity and mechanical properties of polymeric composites through formation of covalent bonds between boron nitride and rubber chains

Hexagonal boron nitride (BN) platelets, also known as white graphite, are often used to improve the thermal conductivities of polymeric matrices. Due to the poor interfacial compatibility between BN platelets and polymeric matrices, in this study, polyrhodanine (PRd) was used to modify BN platelets and prepared functionalized BN-PRd platelets, thereby enhancing the interfacial interaction between the thermal conductive filler and polymeric matrix. Then, BN-PRd platelets were dispersed into the nitrile butadiene rubber (NBR) matrix to yield high thermally conductive composites. The presence of N? C═S groups in PRd allowed the combination of PRd and NBR chains containing stable covalent bonds via vulcanization reaction. The thermal conductivity of the as-prepared 30 vol% BN-PRd/NBR composite reached 0.40 W/mK, representing an increment of 135% over pure NBR (0.17 W/mK). In addition, the largest tensile strength of NBR composite containing 30 vol% BN-PRd platelets was 880% times of pure NBR. The 30 vol% BN-PRd/NBR composite also displayed a relatively high dielectric constant (9.35 at 100 Hz) and a low dielectric loss tangent value (0.07 at 100 Hz), indicating their usefulness as dielectric flexible materials of microelectronics. In sum, the simplicity and good efficiency of formation of covalent bonds between boron nitride and rubber chains look very promising for large-scale industrial production of high thermally conductive composites.