基于SRAM的通用存算一体架构平台在物联网中的应用

最近,存算一体(IMC)架构引起了广泛关注,并被认为有望成为突破冯诺依曼瓶颈的新型计算机架构,特别是在数据密集型(data-intensive)计算中能够带来显著的性能和功耗优势.其中,基于SRAM的IMC架构方案也被大量研究与应用.该文在一款基于SRAM的通用存算一体架构平台——DM-IMCA的基础上,探索IMC架构在物联网领域中的应用价值.具体来说,该文选取了物联网中包括信息安全、二值神经网络和图像处理在内的多个轻量级数据密集型应用,对算法进行分析或拆分,并将关键算法映射到DM-IMCA中的SRAM中,以达到加速应用计算的目的.实验结果显示,与基于传统冯诺依曼架构的基准系统相比,利用DM-IMCA来实现物联网中的轻量级计算密集型应用,可获得高达24倍的计算加速比.     

基于大数据深度学习的输电塔结构抗风易损性评估

为了有效预测输电网系统在强风作用下的响应，并开展高效精准的性能评估，文章提出了基于深度学习模型的风致易损性评估框架。以某具备健康监测系统的输电塔结构为例，首先对监测数据进行清洗和重构，通过大数据深度学习建立荷载输入和响应输出的等效映射模型，然后通过数值模拟生成灾害强度均匀的风场数据并由深度学习模型预测输电塔关键杆件响应，计算不同性能水准的易损性曲线。研究结果表明，经训练的深度学习模型可以涵盖实际工程中存在的各类不确定性因素，有效映射复杂风环境下输电塔结构的动力响应。提出的框架方法可以避免单纯通过数值模型制备大量动态响应数据，更高效地进行输电网系统风致易损性评估。     

盾构隧道施工物料无人运输发展现状及关键技术

盾构隧道施工物料无人运输有助于提升运输效率、降低运输成本、减少运输安全事故率。与公路汽车无人驾驶、港口车辆无人驾驶、轨道交通无人驾驶等逐步成熟的无人驾驶技术不同，在隧道场景下实施施工物料无人运输存在运输物料种类繁多、运输调度困难、地下定位信号拒止、狭窄车道频繁会车、行车路面工况复杂、地上地下联动响应慢等诸多挑战。从实现盾构隧道施工物料无人运输的关键问题分析入手，综述当前盾构隧道施工物料运输方式、运输需求、无人运输发展现状及存在的挑战，提出盾构隧道施工物料无人运输的五大关键技术：多种类物料智能货控能效管理技术，无人化垂直装卸门机控制技术，多传感器融合同步定位与地图构建技术，隧道复杂环境路径规划与自主避障技术，隧-地一体化联动响应的高可靠性无线通信技术。     

Cornus officinalis vinegar reduces body weight and attenuates hepatic steatosis in mouse model of nonalcoholic fatty liver disease

This study aimed to determine the main bioactive components of Cornus officinalis vinegar (COV) and assess the effects of COV on the body weight (BW) and hepatic steatosis in a nonalcoholic fatty liver disease (NAFLD) mouse model. Seven-week-old KM female mice were divided into five treatment groups: (1) Normal control (NC) group, (2) high fat diet (HFD) group, (3) low concentration treatment group (3.5% COV), (4) medium concentration treatment group (5.0% COV), and (5) high concentration treatment group (6.5% COV). Mice in the NC group were fed with a normal chow diet, and those in the other four groups were fed with a HFD known for causing obesity for 10 weeks. Then, mice in the three COV treatment groups were orally administered with COV once a day for 6 weeks. Results showed that the contents of loganin and morroniside in COV reached 16.82 and 51.17 µg/ml, respectively, and COV also contained multiple organic acids. COV significantly reduced BW, abdominal fat weight, liver weight, and the levels of glucose, triglyceride, and low-density lipoprotein cholesterol of serum and increased the levels of high-density lipoprotein cholesterol of serum (p < 0.05). COV also improved the liver function and anti-oxidant activity of liver (p < 0.05). COV treatments increased the interleukin-10 expression and reduced the tumor necrosis factor-α expression in the liver tissue of NAFLD mice (p < 0.05). Histopathological observation revealed that COV suppressed hepatic lipid accumulation and steatosis. The results suggest that COV may contribute to the alleviation of NAFLD and obesity.     

A comprehensive review of image analysis methods for microorganism counting: from classical image processing to deep learning approaches

Microorganisms such as bacteria and fungi play essential roles in many application fields, like biotechnique, medical technique and industrial domain. Microorganism counting techniques are crucial in microorganism analysis, helping biologists and related researchers quantitatively analyze the microorganisms and calculate their characteristics, such as biomass concentration and biological activity. However, traditional microorganism manual counting methods, such as plate counting method, hemocytometry and turbidimetry, are time-consuming, subjective and need complex operations, which are difficult to be applied in large-scale applications. In order to improve this situation, image analysis is applied for microorganism counting since the 1980s, which consists of digital image processing, image segmentation, image classification and suchlike. Image analysis-based microorganism counting methods are efficient comparing with traditional plate counting methods. In this article, we have studied the development of microorganism counting methods using digital image analysis. Firstly, the microorganisms are grouped as bacteria and other microorganisms. Then, the related articles are summarized based on image segmentation methods. Each part of the article is reviewed by methodologies. Moreover, commonly used image processing methods for microorganism counting are summarized and analyzed to find common technological points. More than 144 papers are outlined in this article. In conclusion, this paper provides new ideas for the future development trend of microorganism counting, and provides systematic suggestions for implementing integrated microorganism counting systems in the future. Researchers in other fields can refer to the techniques analyzed in this paper.

     

Gelling behavior of MgAl2O4 slurries with a common dispersant

MgAl₂O₄ transparent ceramics were shaped by a commonly used polyacrylic acid (PAA), which acted as both dispersant and gelling agent. The spinel slurries were prepared by ball-milling MgAl₂O₄ powder, PAA, and water in an attrition mill. The gelling of slurries happened at room temperature in air atmosphere without any other organic additive. The gelling mechanism was the formation of chelates between Mg²⁺ and carboxyl groups (-COO⁻) of PAA. The frequency-based testing method was applied to investigate the gelling process of the as-prepared slurry. In addition, a novel in situ characterization method based on a modified indentation testing was invented to better understand the strengthening of the wet green body with time and to guide when demolding could be carried out. After sintering, transparent MgAl₂O₄ ceramics with high in-line transmittance were resulted.     

Deep learning-based symbol detection algorithm in IMDD-OOFDM system

In the current research on intensity-modulation and direct-detection optical orthogonal frequency division multiplexing ( IMDD-OOFDM ) system, effective channel compensation is a key factor to improve system performance. In order to improve the efficiency of channel compensation, a deep learning-based symbol detection algorithm is proposed in this paper for IMDD-OOFDM system. Firstly, a high-speed data streams symbol synchronization algorithm based on a training sequence is used to ensure accurate symbol synchronization. Then the traditional channel estimation and channel compensation are replaced by an echo state network (ESN) to restore the transmitted signal. Finally, we collect the data from the system experiment and calculate the signal-to-noise ratio (SNR). The analysis of the SNR optimized by the ESN proves that the ESN-based symbol detection algorithm is effective in compensating nonlinear distortion.     

双功能催化剂上四氢萘加氢裂化生成单环芳烃的反应网络研究

Conversion of LCO (light cycle oil) to BTX (benzene, toluene, and xylene) is an economically valuable method for refineries. However, this approach still faces difficulties as the main reactions are not clearly understood. Here we study the detailed hydrocracking pathway of typical reactants, 1-methylnaphthalene and tetralin, through molecular simulations and experiments to improve our understanding of the conversion process of LCO to BTX. Molecular simulations demonstrate that the rate-determining step is the isomerization pathway of six-membered ring to five-membered ring in tetralin as its activation energy (ΔEa) is the highest among all the reactions and the order of ΔEa of reactions is isomerization > ring-opening ≈ side-chain cleavage. The results of experiments show that with the increase in reaction depth, i.e., through a high temperature (350 – 370 °C) and low LHSV (4.5 – 6.0 h−1), isomerization, ring-opening, and side-chain  cleavage reactions occurred, thus improving the selectivity and yield of alkyl aromatics.