Investigation of fabrication of gas diffusion substrate for proton exchange membrane fuel cells

The gas diffusion substrate (GDS) is essential in the proton exchange membrane fuel cells. Its fabrication techniques affect the performance significantly and are worthy of investigation. In this study, a manufacturing process of the GDS is proposed to understand the formation process of GDS and promote its structure and performance more pertinently. Different states during the preparation process, raw carbon paper, pre-curing, curing, carbonation, and graphitization, are characterized and measured. Experimental and numerical methods are employed to determine the relationships between microstructure, transport, and mechanical performance variation with the fabricating processes. The results show that its porosity, average pore size, and effective diffusivity decrease first and increase after curing. These parameters after graphitization are lower than that of the carbon paper (CP). The electrical resistivity increases dramatically while pre-curing and decreases gradually after curing, carbonation, and graphitization, and it is much reduced after graphitization. Moreover, mechanical measurement results show that both the picks of tensile strength and flexural modulus occur after curing. Its tensile strength shows little change after graphitization compared to the initial paper's. In contrast, the flexural modulus is improved significantly.     

One-step to prepare high-performance gas diffusion layer (GDL) with three different functional layers for proton exchange membrane fuel cells (PEMFCs)

Gas diffusion layer (GDL) is one of the most important components of fuel cells. In order to improve the fuel cell performance, GDL has developed from single layer to dual layers, and then to multiple layers. However, dual or multi layers in GDL are usually prepared by layer-by-layer methods, which cost too much time, energy, and resources. In this work, we successfully developed a facile one-step method to prepare a GDL with three functional layers by utilizing the different sedimentation rates and filtration rates of short carbon fiber (CF) and carbon nanotube (CNT). The treatment temperature for this GDL is much lower than that of traditional method. The thickness of the GDL can be effectively controlled from as thin as 50 μm to more than 200 μm by simply adjusting the content of CF. The GDL with high flexibility is suitable to develop high performance flexible electronics. The fuel cell with the GDL has the maximum power density 1021 mW cm^?2, which shows 19% improvement comparing to the conventional one. Therefore, this work breaks the traditional concept that GDL for fuel cells only can be prepared by very complex and high-cost procedure.     

Probabilistic learning on manifolds (PLoM) with partition

The probabilistic learning on manifolds (PLoM) introduced in 2016 has solved difficult supervised problems for the “small data” limit where the number N of points in the training set is small. Many extensions have since been proposed, making it possible to deal with increasingly complex cases. However, the performance limit has been observed and explained for applications for which N is very small and for which the dimension of the diffusion-map basis is close to N. For these cases, we propose a novel extension based on the introduction of a partition in independent random vectors. We take advantage of this development to present improvements of the PLoM such as a simplified algorithm for constructing the diffusion-map basis and a new mathematical result for quantifying the concentration of the probability measure in terms of a probability upper bound. The analysis of the efficiency of this extension is presented through two applications.     

机器翻译译文质量估计综述

机器翻译译文质量估计(Quality Estimation,QE)是指在不需要人工参考译文的条件下,估计机器翻译系统产生的译文的质量,对机器翻译研究和应用具有很重要的价值。机器翻译译文质量估计经过最近几年的发展,取得了丰富的研究成果。该文首先介绍了机器翻译译文质量估计的背景与意义;然后详细介绍了句子级QE、单词级QE、文档级QE的具体任务目标、评价指标等内容,进一步概括了QE方法发展的三个阶段: 基于特征工程和机器学习的QE方法阶段,基于深度学习的QE方法阶段,融入预训练模型的QE方法阶段,并介绍了每一阶段中的代表性研究工作;最后分析了目前的研究现状及不足,并对未来QE方法的研究及发展方向进行了展望。     

快速近似计算Shapley值的归因解释方法

Shapley值归因解释方法虽然能更准确量化解释结果, 但过高的计算复杂度严重影响了该方法的实用性. 本文引入KD树重新整理待解释模型的预测数据, 通过在KD树上插入虚节点, 使之满足TreeSHAP算法的使用条件, 在此基础上提出了KDSHAP方法. 该方法解除了TreeSHAP算法仅能解释树结构模型的限制, 将该算法计算Shapley值的高效性放宽到对所有的黑盒模型的解释中, 同时保证了计算准确度. 通过实验对比分析, KDSHAP方法的可靠性, 以及在解释高维输入模型时的适用性.     

Automated prediction system of household energy consumption in cities using web crawler and optimized artificial intelligence

The supply of electrical energy is critical to convenient and comfortable living. However, people consume a large amount of energy, contributing to an energy crisis and global warming, and damaging some ecological cycles. Residential electricity consumption has greater elasticity than industrial and business consumption; it therefore has high energy-saving potential. This work establishes an automated platform, which provides information about residential electricity consumption in each city in Taiwan. Machine learning was used to forecast future residential electricity demand. A nature-inspired optimization method was applied to enhance the accuracy of the best machine learner, yielding an even better hybrid ensemble model. Performance measures indicate that the resulting model is accurate and provides effective information for reference. An automatic web-based system based on the model was combined with a web crawler and scheduled to run automatically to provide information on monthly residential electricity consumption in each county and city. By providing energy consumption information across the country, power providers and government can discuss policy and set different goals for energy use. The results of this study can facilitate the early implementation of energy-saving and carbon emission-reducing in cities and aid utility companies in establishing energy conservation guidelines.     

Prediction of fish quality level with machine learning

In this study, sea bream, sea bass, anchovy and trout were captured and recorded using a digital camera during refrigerated storage for 7 days. In addition, their total viable counts (TVC) were determined on a daily basis. Based on the TVC, each fish was classified as ‘fresh’ when it was <5 log cfu per g, and as ‘not fresh’ when it was >7 log cfu per g. They were uploaded on a web-based machine learning software called Teachable Machine (TM), which was trained about the pupils and heads of the fish. In addition, images of each species from different angles were uploaded to the software in order to ensure the recognition of fish species by TM. The data of the study indicated that the TM was able to distinguish fish species with high accuracy rates and achieved over 86% success in estimating the freshness of the fish species tested.     

Sulfate- and pH-driven metabolic flexibility in sugarcane vinasse dark fermentation stimulates biohydrogen evolution,sulfidogenesis or homoacetogenesis

Dark fermentation of sugarcane vinasse can be used as a “cleaning” step to remove sulfate prior to methanogenesis because sulfidogenic conditions can be successfully established in parallel with biohydrogen production. Using a 2² central composite rotational design (CCRD) and response surface methodology (RSM), this study assessed the impacts of bicarbonate and sulfate availability on the establishment of sulfidogenesis in the thermophilic (55 °C) fermentation of vinasse in batch reactors, equally assessing the impacts on biohydrogen evolution. CCRD-RSM results indicated the favoring of biohydrogen production at the lowest sulfate and bicarbonate concentrations, whilst the opposite was observed for sulfidogenesis. Glycerol, lactate, and hydrogen were the preferential electron donors utilized by sulfate-reducing bacteria (SRB), whilst ethanol was markedly consumed only at high sulfate concentrations. SRB were inhibited by sodium when dosing excess NaHCO₃ and Na₂SO₄. Complementary tests revealed maximum biohydrogen production (2.40 mmol) out of the CCRD, at pH exceeding 7.5 with no interference of sulfidogenesis. Non-efficient biohydrogen production was observed at low pH (<5.0; ～1.90 mmol) because the uptake of lactate was inhibited. Meanwhile, homoacetogenesis was established under intermediate pH range (5.5–6.5), as revealed by the accumulation of acetate (up to 2.5 g L^?1). 16S rRNA gene amplicon sequencing further revealed the genera Thermoanaerobacterium/Pseudoclostridium, Desulfotomaculum/Desulfohalotomaculum and Sporomusaceae/Moorella as the main biohydrogen-producing, sulfate-removing and biohydrogen-consuming (homoacetogens) microbial groups, respectively. Hence, using a single inoculum source, vinasse may provide a butyrate-rich (along with biohydrogen-rich biogas) or a sulfate-free and acetate-rich fermented effluent, depending mainly on proper pH control.     

Tunable BaxSr1-xTiO3 nanoparticles induced high energy storage density in layer-structured asymmetric polymer-based nanocomposites

In order to solve the problem of low charging and discharging energy density of dielectric capacitors, the structure design of layered polymer matrix composites is carried out in this paper. Ba_0.7Sr_0.3TiO₃, Ba_0.8Sr_0.2TiO₃ and Ba_0.9Sr_0.1TiO₃ nanoparticles were successfully prepared by the oxalate coprecipitation method. The surface of Ba_xSr_1-xTiO₃ was successfully coated with dopamine, which promoted the dispersion of the polymer matrix of the ceramic powder. Monolayer Ba_xSr_1-xTiO₃/PVDF composites containing Ba_xSr_1-xTiO₃ with different Ba/Sr ratios were successfully prepared by the casting method. Three-layer asymmetric composites with different fillers were successfully prepared by layer-by-layer casting. The phase and microstructure of the as-prepared materials were analyzed by XRD and SEM. The dielectric, electrical conductivity, ferroelectric and energy storage properties of the composites were tested. The effects and laws of the design of the three-layer asymmetric structure on the dielectric properties and energy storage properties of the layered composites are mainly studied. When the structure of the three-layer asymmetric composite is 1-2-3, the breakdown field strength reaches 330 kV/mm, the discharge energy density reaches 8.51 J/cm³, and the charge-discharge efficiency is 67%. This work demonstrates that layered composites with asymmetric properties can facilitate the development of electrical energy storage.     

毫米波三维异质异构集成技术的进展与应用

三维异质异构集成技术是实现电子信息系统向着微型化、高效能、高整合、低功耗及低成本方向发展的最重要方法,也是决定信息化平台中微电子和微纳系统领域未来发展的一项核心高技术。文章详细介绍了毫米波频段三维异质异构集成技术的优势、近年来的发展趋势以及面临的挑战。利用硅基MEMS 光敏复合薄膜多层布线工艺可实现异质芯片的低损耗互连,同时三维集成高性能封装滤波器、高辐射效率封装天线等无源元件,还能很好地处理布线间的电磁兼容和芯片间的屏蔽问题。最后介绍了一款新型毫米波三维异质异构集成雷达及其在远距离生命体征探测方面的应用。