Research on Crop Irrigation Schedules Under Deficit Irrigation—A Meta-analysis

Water Resources Management - Increasing water consumption in agriculture due to global climate change has posed considerable challenges to food security, thus improving the efficiency of water...     

用于疏浚工程泥浆密度测量的电阻层析成像技术

江河湖海疏浚工程对于拓展和保障航运安全、促进生态恢复、维护/改善远海岛礁环境具有不可或缺的作用.目前,在绞吸和耙吸疏浚船上,多采用射线密度计及电磁流量计对输泥密度、产量等进行在线计量,并作为实现疏浚系统自动控制的重要参量.由于射线密度计潜在的人员健康、公共安全和环保风险,在一些国家和地区已建立了放射源的有序退出机制.电阻层析成像使用安装于管壁的电极阵列获得管道内流体的电导率分布,进而推导出管道内的泥浆密度,具有无辐射、全场测量及实时可视化的优点.电阻层析成像测量系统经过特殊设计,可适应从内陆淡水到海水的巨大电导率变化.初步测试表明,电阻层析成像测量数据与射线密度计具有良好一致性,具备了作为在线泥浆密度测量的可行性.此外,ERT可提供实时的流动成像,对于提高对疏浚管道监测能力和降低堵管概率具有重要意义.     

 基于混合模型的优化低频低压减载方法研究

文中以低频低压情况下保证电压稳定的前提下减载为研究内容,提出了一种基于混合模型的优化低频低压减载方法。该混合模型将遗传算法与人工神经网络相结合,在保证电压稳定的情况下选择合适的节点进行负荷切除。实验结果表明,该方法可以在低频低压情况下使负荷减载量达到最小。     

35kV变压器故障气相色谱分析

对于运行中的变压器,气相色谱分析法可实现其内部故障的诊断,并提供有效的运维建议.应用气相色谱分析法分析一起35 kV变压器故障,为气相色谱分析法的进一步推广应用提供有力的技术支撑.     

"为你写诗":面向中国古典诗歌的可视化交互创作系统

中国古典诗歌具有工整的句式和悠扬的韵律,这些特点对创作者提出了较高的要求.现有的诗歌生成系统大多着眼于生成模型的效果,用户的参与度和发挥空间有限.基于古诗文网、THUNLP-AIPoet等开源数据,设计并实现了一个面向中国古典诗歌的可视化交互创作系统"为你写诗".其自定义视图支持自定义句式情感;"挥毫泼墨"视图自动生成诗歌作为创作基础,并且可视化诗歌文本的韵律情感等多维度属性;修改辅助视图在创作过程中提供改进建议和候选推荐.通过实例分析和用户调研对系统视图的功能和创作结果的连贯性、情感和韵律进行评价,结果表明了它的有效性和实用性.     

一类企业竞争模型的动力学分析

建立了具有产出水平下临界点的企业竞争型模型,研究了该系统的平衡点的存在性及稳定性,并通过数值模拟的方法验证了研究结果,最后给出了模型的理论解释。     

Phi zeolite synthesized by template-free method for selective catalytic reduction of NO

Nitrogen oxide (NO_x) is one of the major air pollutants. Using a copper-based zeolite with a chabazite structure (CHA) as a catalyst, the selective catalytic reduction (SCR) technology can effectively remove NO_x. This work introduces a low silica CHA type zeolite with structural defects (zeolite Phi, with Si/Al of 4.7). The zeolite Phi is synthesized through a hydrothermal method without adding any template, which is low-cost and environment-friendly. The Cu-exchanged Phi is abundant of surface acidity and isolated Cu²⁺, showing a superior low-temperature activity, a wide work temperature window and a good hydrothermal stability. The presence of Na or Mg decreases the surface acidity and isolated Cu²⁺. The hydrothermally aged Na,Cu/Phi and Mg,Cu/Phi present different levels of framework collapse, which correspondingly induces catalyst deactivation.     

High-rate Decoupled Water Electrolysis System Integrated with α-MoO3 as a Redox Mediator with Fast Anhydrous Proton Kinetics

Hydrogen is a promising alternative to fossil fuels that can reduce greenhouse gas emissions. Decoupled water electrolysis system using a reversible proton storage redox mediator, where the oxygen evolution reaction and hydrogen evolution reaction are separated in time and space, is an effective approach to producing hydrogen gas with high purity, high flexibility, and low cost. To realize fast hydrogen production in such a system, a redox mediator capable of releasing protons rapidly is required. Herein, α-MoO₃, with an ultrafast proton transfer property that can be explained by a dense hydrogen bond network in the lattice oxygen arrays of HxMoO₃, is examined as a high-rate redox mediator for fast hydrogen production in acidic electrolytes. The α-MoO₃ redox mediator shows both a large capacity of 204 mAh g⁻¹ and fast hydrogen production at a current rate of 10 A cm⁻²(≈153 A g⁻¹), outperforming most of the previously reported solid-state redox mediators.     

数字孪生贾鲁河在工程施工期防洪度汛中的应用

为满足贾鲁河综合治理工程施工期防洪度汛精细化管理的需求,确保工程安全的同时能够充分发挥流域防洪效益,数字孪生技术被应用于在建水利工程施工度汛精准风险分析、预警和防汛预案制定中。数字孪生贾鲁河以自建采集L3级和共享L1和L2级数据为“算据”底板,以模型平台和知识平台为“算法”支撑,以信息化基础设施为“算力”依托,搭建具有预报、预警、预演、预案等“四预”功能的流域防洪应用。在贾鲁河流域基础资料收集及数字化场景搭建和模型构建基础上,以流域降雨量级超100年一遇、约50年一遇、约20年一遇设置了三类预演情景模拟洪水,统计施工期工程存在的漫堤风险、淹没风险、冲刷风险、洪水倒灌风险、桥梁阻水风险、堤外积水风险等六类风险并制定防洪度汛措施。数字孪生贾鲁河确保了工程2022年安全度汛,总结了河道治理工程施工期洪水风险类型和定量评估方法,形成了在建工程复杂环境条件下的地形勘测、洪水模拟、风险预警、应对措施制定的完整解决方案。     

Weakly supervised temporal action localization with proxy metric modeling

Temporal localization is crucial for action video recognition. Since the manual annotations are expensive and time-consuming in videos, temporal localization with weak video-level labels is challenging but indispensable. In this paper, we propose a weakly-supervised temporal action localization approach in untrimmed videos. To settle this issue, we train the model based on the proxies of each action class. The proxies are used to measure the distances between action segments and different original action features. We use a proxy-based metric to cluster the same actions together and separate actions from backgrounds. Compared with state-of-the-art methods, our method achieved competitive results on the THUMOS14 and ActivityNet1.2 datasets.