Subjects: Computer Science >> Natural Language Understanding and Machine Translation submitted time 2024-01-04
Abstract： This paper explores potential directions for the application of large language models in the field ofYi studies, starting from the characteristics of Yi studies and the main scenarios for applying largelanguage models. It aims to investigate the potential role of large language models in Yi studies,highlighting specific applications of these models in Yi divination as an example. Yi studies havegarnered attention due to their extensive cultural significance, profound theoretical frameworks, andpractical applications, characterized by high specialization and a steep learning curve. Large languagemodels have achieved significant breakthroughs in natural language processing, knowledge graphconstruction, with major applications including machine translation, text summarization, intelligentquestion answering, and semantic understanding. However, the utilization of large language modelsin Yi studies remains largely underexplored. Therefore, this paper proposes potential directions forthe application of large language models in Yi studies, such as Yi literature analysis and interpretationof Yi divination results. Finally, by elaborating on a detailed application instance of large languagemodels in Yi divination, this paper reveals the potential of large language models in the field of Yistudies. A comprehensive understanding of the characteristics of Yi studies and extensive explorationof the application scenarios of large language models will contribute to fostering a broader applicationof these models in Yi studies, enhancing research efficiency, and promoting cultural inheritance.
Peer Review Status:
Awaiting Review
Subjects: Dynamic and Electric Engineering >> Electrical Engineering submitted time 2019-03-05 Cooperative journals: 《电气工程学报》
Abstract：主要介绍了几种抑制逆变器共模电压（CMV）的方法，其中一类是基于空 间矢量调制策略优化的方案，主要包括零矢量替代（AZSPWM）、最近非零矢量合成 （NSPWM）、虚拟空间矢量调制（VSVM）等策略；另一类是改变硬件电路的拓扑结构 抑制共模电压，主要是通过利用三相四桥臂拓扑结构完成对共模电压的抑制。对上述 两类方法进行相关的仿真与分析，为共模电压抑制方案的研究提供参考。
Subjects: Biology >> Bioengineering submitted time 2019-03-05 Cooperative journals: 《中国生物工程杂志》
Abstract： Objective: A new material with ultra-large surface area named nano-forest is prepared by Micro-Electro-Mechanical System(MEMS) processing technology. Based on this material, a new microfluidic chip for point-of-care test with simple operation, rapid detection and high sensitivity is created. Methods: The fabrication of nano-forests in micro-channel on quartz substrate mainly includes: cleaning and drying of quartz substrate; spinning polyimide(PI) coating; re-spinning phenolic resin photoresist on PI coating; photolithography to expose the channel; treating the PI layer with oxygen plasma and argon plasma to synthesize nano-fiber forests structure; nano-fiber-quartz nanoforests are realized by using nano-fiber forests as nanomasks in anisotropic etching of quartz by using reactive ion etching (RIE); the micro-channel with nano-forests structure inside is achieved after removing upper nanofiber forests structure and phenolic resin photoresist coating.The height, width, density and specific surface area of nano-forest are studied and analyzed by scanning electron microscope(SEM). Optical properties are tested by ultraviolet-visible spectrophotometer. The driving force is characterized by the flow rate of PBS solution.The sensitization effect is evaluated by saturated fluorescence test through antibody and AbFluor 680 dye-labeled secondary antibody. The sample pad, bond pad, micro-channel with nano-forests structure, nitrocellulose membrane and absorbent material are assembled on PMMA substrate in sequence, which is the microfluidic chip. The chip based on the sandwich format with a polyclonal antibody and a AbFluor 680 dye-labeled secondary antibody is used to detect ricin toxin(RT). Results: The scanning electron microscope shows that the nanofiber forests structure is formed on quartz substrate after oxygen plasma and argon plasma bombardment. The single nanofiber is upright on the substrate with a diameter of about 50-100 nm, a height of 1.8 um and a density of about 20/μm2. The quartz nano-forests structure can be obtained after RIE with nano-fibre forests structure as mask and resist removal. The single structure is shaped like a cone. The diameter of the cone bottom is about 100-200 nm, the height is about 1.0 um, the density is about 10/μm2, and the surface area to bottom area is more than 5:1. Self-driven test provides information of the flow rate of PBS is to be about 5 mm/s in the micro-channel on the basis of nano-forests structure. The transmittance of the channel is 89.5%at 680 nm wavelength. It shows that the channel has good transmittance, which makes the loss of excitation light or emission light much less, and is conducive to the sensor capturing more signals. With same surface modification, the planar quartz structure has shortcomings of short lasting effect time and low saturation fluorescence intensity. To the contrary, nano-forests structure with ultra-large surface area has a good sensitization effect in the test. RT can be detected sensitively based on the significantly fluorescent intensity.The linear range of detection is from 10 pg/mL to 6250 pg/mL and the limit of detection (LOD) is lower than 10 pg/mL. Conclusion: The nano-forests structure with good optical properties reduces the requirements of sensor and also makes the choice of fluorescent dyes wider.The three-dimensional structure of the nano-forest has an ultra-large surface area, which increases the amount of antibody compared to the planar structure, and thus improves the sensitivity of detection greatly. Compared with the immunochromatographic test strip, the microfluidic chip has an advantage of high sensitivity, thus the quantitative analysis can be realized within a certain range. Most microfluidic chips require complex equipments to provide driving force, which will make them costly and bulky. Driven by the capillary force, the chip with nano-forests structure inside makes the detection simple and fast. Combined with the miniaturized detection terminal, the platform can be miniaturized, portable, and automated, achieving the goal of simple, fast and efficient analysis. These characteristics make the chip an ideal candidate for the development of rapid detection methods.
Subjects: Biology >> Botany >> Plant ecology, plant geography submitted time 2016-05-03
Abstract： Aims Forest carbon storage in Inner Mongolia plays a significant role in national terrestrial carbon budget due to its largest forest area in China. Our objectives were to estimate the carbon storage in the forest ecosystems in Inner Mongolia, and to reveal its spatial pattern. Methods In the study, field survey and sampling were conducted at 137 sites, which were nearly evenly distributed in most forest types in the study region. At each site, the ecosystem carbon density at each site was estimated thorough sampling and measuring pools of soil (0–100 cm) and vegetation, including biomass of tree, grass, shrub, and litter. Regional carbon storage was further calculated with the estimated carbon density for each forest type. Important findings Carbon storage of vegetation layer in forests in Inner Mongolia was 787.8 Tg C, with the biomass of tree, litter, herbaceous and shrub account for 93.5%, 3.0%, 2.7% and 0.8%, respectively. Carbon density of vegetation layer was 40.4 t?hm–2, with 35.6 t?hm–2 in trees, 2.9 t?hm–2 in litter, 1.2 t?hm–2 in herbaceous and 0.6 t?hm–2 in shrubs. In comparison, carbon storage of soil layer in forests in Inner Mongolia was 2449.6 Tg C, with 79.8% distributed in the first 30 cm. Carbon density of soil layer was 144.4 t?hm–2. Carbon storage of forest ecosystem in Inner Mongolia was 3237.4 Tg C, with vegetation and soil accounting for 24.3% and 75.7%, respectively. Carbon density of forest ecosystem in Inner Mongolia was 184.5 t?hm–2. Carbon density of soil layer was positively correlated with that of vegetation layer. Spatially, both carbon storage and carbon density were higher in the eastern area, where the climate is more humid. Forest reserve construction and artificial afforestation can significantly improve the capacity of regional carbon sink.
Peer Review Status:Awaiting Review
Subjects: Biology >> Botany >> Plant ecology, plant geography submitted time 2016-05-03
Abstract： Aims Sparse Ulmus pumila forest is an intrazonal vegetation in Onqin Daga Sandy Land, Populus simonii has been widely planted for windbreak and sand dune stabilization. Our objective was to compare the differences in carbon (C) density of these two forests and their relationships with stand age. Methods We measured the C content of tree organs (leaf, twig, stem, and root), herb layers(above ground vegetation and below ground root) and soil layers (up to 100 cm) in sparse Ulmus pumila forests and Populus simonii plantations of different stand ages, and then computed C density and their proportions in total ecosystem carbon density. In addition, we illustrated the variation of tree layer, soil layer and total ecosystem carbon density with stand age, and then estimated the C sequestration rates for these two ecosystems by the space-for-time substitution approach. Important findings The average C contents of tree layer and soil layer for sparse Ulmus pumila forests were lower than those for Populus simonii plantations. The total C density of sparse Ulmus pumila forests was half of that of Populus simonii plantations. The carbon density of soil and tree layers accounted for more than 98% of ecosystem C density in both two forests. Irrespective of forest type, the C density ratios of soil to vegetation decreased with stand age. This ratio was 1.66 for sparse Ulmus pumila forests and 1.87 for Populus simonii plantations, respectively when they were over-matured. The C density of tree layer, soil layer, and total ecosystem in both forests increased along forest development. There were significant linear positive correlations between tree layer’s C density and stand age in both forests and between the total ecosystem C density of Sparse Ulmus pumila forests and stand age. The C sequestration rate of tree layer was 5-fold higher in Populus simonii plantation than in sparse Ulmus pumila forest. The C sequestration rate of ecosystem was 0.81 Mg C hm–2 a–1 for sparse Ulmus pumila forest and 5.35 Mg C hm–2 a–1 for Populus simonii plantation. These findings have implications for C stock estimation of sandy land forest ecosystems and policy-making of ecological restoration and C sink enhancement in the studied area.
Peer Review Status:Awaiting Review
Hits
Hits
Downloads
Comment