Factor analysis using delta-rule wake-sleep learning

We describe a linear network that models correlations between real-valued visible variables using one or more real-valued hidden variables-a factor analysis model. This model can be seen as a linear version of the Helmholtz machine, and its parameters can be learned using the wake-sleep method, in which learning of the primary generative model is assisted by a recognition model, whose role is to fill in the values of hidden variables based on the values of visible variables. The generative and recognition models are jointly learned in wake and sleep phases, using just the delta rule. This learning procedure is comparable in simplicity to Hebbian learning, which produces a somewhat different representation of correlations in terms of principal components. We argue that the simplicity of wake-sleep learning makes factor analysis a plausible alternative to Hebbian learning as a model of activity-dependent cortical plasticity.     

The dispersion of solute in turbulent pipe flow of non-newtonian fluids

The dispersion of solute in turbulent pipe flow of power law fluids is investigated theoretically. Results in terms of the intensity of dispersion are compared for the non-Newtonian and purely viscous Newtonian fluids over a wide range of Schmidt numbers. The computed results indicate that the intensity of dispersion at a constant Schmidt and Reynolds number increases with increasing flow behavior index.     

湍流液环柱塞流动分析

本文应用Spalding的精确内层律,构造了湍流液环柱塞流动的“双层边界”模型,得出了可用于计算偏心摩阻的Λ-Re图,所得结果经实验初步验证。     

En Route to High-Density Chiral Single-walled Carbon Nanotube Arrays using Solid Trojan Catalysts

Solid catalyst is widely recognized as an effective strategy to control the chirality of single-walled carbon nanotubes (SWNTs). However, it is still not compatible with high density in horizontal arrays. “Trojan” catalysts strategy is one of the most effective methods to realize SWNTs with high density and has great potential in chirality control. Here, the co-realization of high density and chirality controlling for SWNTs in a low-temperature growth process is reported based on the developed solid “Trojan” catalyst. High temperature “Trojan” catalyst formation process provides sufficient catalyst number to acquire high density. These liquid “Trojan” catalysts are cooled to solid state by adopting low growth temperature (540 °C), which can be good template to realize the chirality controlling of SWNTs with exposing six-fold symmetry face, (111). Finally, (9, 6) and (13, 1) SWNTs enriched horizontal array with the purity of ≈90% and density of 4 tubes µm⁻¹ is realized. The comparison between the distribution of initial catalysts and the density of as-grown tubes indicates no sacrificing on catalysts number to improve chirality selectivity. This work opens a new avenue on the catalyst's design and chirality controlling in SWNTs growth.     

Improving Photovoltaic Performance of Non-Fullerene Polymer Solar Cells Enables by Fine-Tuning Blend Microstructure via Binary Solvent Mixtures

Studies of the relationship between blend microstructure and photovoltaic performance are becoming more common, which is a prerequisite for rationally improving device performance. Non-fullerene acceptors usually have planar backbone conformation and strong intermolecular packing, normally resulting in excessive phase separation. Herein, an effective co-solvent blending strategy to turn the molecular organization of a chlorinated small molecule acceptor Y6-2Cl and phase separation of the corresponding active layer with PM6 as donor is demonstrated. The in situ photoluminescence measurements and relevant morphological characterizations illustrate that the film-forming process is fine-turned when using the mixtures of chloroform (CF) and chlorobenzene (CB) solvents, and the blend showed high domain purity with suitable phase-separated networks. Thus, better exciton dissociation and charge generation, more balanced charge transport, and less recombination loss are obtained in the co-solvent blade-coated devices. As a result, a maximum power conversion efficiency (PCE) of 16.17% is achieved, which is much higher than those of CF- and CB-bladed devices (14.08% and 11.44%, respectively). Of note is that the use of this co-solvent approach in the other two high-performance photovoltaic systems is also confirmed, demonstrating its good generality of employing in the printing organic solar cells.     

Reactions of fuel-nitrogen compounds under conditions of inert pyrolysis

As part of a study of the chemical mechanisms involved in the conversion of fuel-nitrogen compounds to nitric oxide during combustion, fossil fuels and model nitrogen compounds were pyrolysed in helium in a small quartz flow reactor. Hydrogen cyanide was the major nitrogen-containing product obtained in all cases indicating that hydrogen cyanide is formed during the initial pre-flame stages of combustion and is the principal intermediate in the formation of fuel nitric oxide. At a nominal residence time of one second, 50% decomposition of pyrrole, quinoline, benzonitrile and pyridine occurs at 905, 910, 922 and 954 °C, respectively. The fraction of the nitrogen in pyridine that is converted to hydrogen cyanide increases from 40% at 960 °C to 100% at 1100 °C. Benzonitrile produces similar amounts of hydrogen cyanide (49 and 82%). The hydrogen cyanide yields from coals and residual fuel oils increase from the range of 15–25% at 950 °C to 23–42% at 1100 °C. It is not possible to determine from these single-stage experiments if most of the hydrogen cyanide forms in the primary pyrolysis process or in secondary reactions.     

Dispersion of PCB in the environment following an atmospheric release caused by a fire

During a fire at a power plant located in the coastal plain of Israel, PCBs were released to the atmosphere from a ruptured transformer. Since PCBs are probably carcinogenic to humans, this study was performed in order to assess the environmental contamination by PCBs via the atmospheric pathway and the need for remediation measures. The release conditions and the meteorological conditions which prevailed during the fire were analyzed. This provided the input to a Gaussian dispersion model used to estimate the downwind-contaminated sector as well as the location of the maximal concentration within this sector. A sampling plan was then devised and vegetation collected within this sector was analyzed for PCBs. A methodology was developed to convert PCB concentrations in vegetation to concentrations in the atmosphere. It allowed a reconstitution of the PCB source term to the atmosphere from the vegetation measurements. The PCB concentrations were found to be lower than the USEPA decontamination standards. Remediation measures were not needed beyond the plant fence.     

双型钢混凝土转换梁柱节点抗震性能试验研究

以广东某高层建筑为工程背景,针对大跨度转换结构,提出了双型钢混凝土转换梁柱节点的构造组合形式,通过对2个转换节点的竖向和水平荷载作用下的低周反复荷载试验,研究了节点的破坏形态、承载能力、刚度、滞回特性、延性、耗能能力及关键位置钢筋和型钢的应变等性能。试验结果表明：转换梁内置双型钢腹板形成的封闭空间对混凝土有约束作用,提高了节点区混凝土的抗剪能力;双型钢混凝土转换梁柱节点的滞回曲线饱满,极限变形能力较强,承载力较高,刚度、延性和耗能能力均较好;被转换柱与双型钢混凝土梁采用“端板螺栓连接”实现了“在被转换柱底部先出现塑性铰”,达到了“强梁强柱,更强节点”和“强转换层,弱框架层”目的。     

“双碳”目标下中国居民食物消费结构升级及碳排放分析研究

目的 分析2014—2020年间中国居民食物消费结构及碳排放量,为倡导居民可持续消费模式,实现“双碳”目标提供科学支持。方法 以食物消费为对象,并进一步延伸至食物碳排放。综合运用食物碳排放系数和生命周期法,从直接和间接两个方面对我国居民食物消费引起的碳排放量进行全面的测算,探究食物碳排放演变趋势。结果 中国居民食物结构随动物性食物消费量逐年递增,温室气体排放量也有较大幅度增长。分别假设2030、2040、2050年逐渐调整其食物消费结构至膳食指南场景,温室气体排放总量分别在2030、2040、2049年达到峰值,温室气体排放量较现状水平均有一定幅度增加。结论 中国食物消费结构调整带来的碳减排压力较大,并结合上述结果结论,给出合理的政策启示。