Flow Direction Level Traffic Flow Prediction Based on a GCN-LSTM Combined Model

Traffic flow prediction plays an important role in intelligent transportation systems and is of great significance in the applications of traffic control and urban planning.Due to the complexity of road traffic flow data,traffic flow prediction has been one of the challenging tasks to fully exploit the spatiotemporal characteristics of roads to improve prediction accuracy.In this study,a combined flow direction level traffic flow prediction graph convolutional network(GCN)and long short-term memory(LSTM)model based on spatiotemporal characteristics is proposed.First,a GCN model is employed to capture the topological structure of the data graph and extract the spatial features of road networks.Additionally,due to the capability to handle long-term dependencies,the longterm memory is used to predict the time series of traffic flow and extract the time features.The proposed model is evaluated using real-world data,which are obtained from the intersection of Liuquan Road and Zhongrun Avenue in the Zibo High-Tech Zone of China.The results show that the developed combined GCNLSTM flow direction level traffic flow prediction model can perform better than the single models of the LSTM model and GCN model,and the combined ARIMA-LSTM model in traffic flow has a strong spatiotemporal correlation.

Image Filter for Removal of Impulse Noise Based on Directional Derivatives

One of the most common image processing tasks involves the removal of noise from images. Noise can be introduced during image capture, during transmission, or during storage. For design purposes, noise sources are frequently approximated by random variables with a known probability distribution. One common noise model corrupts a signal by introducing impulses. And the surface of the image disturbed by impulse noise displays many peaks or vales. According to the characteristic of impulse noise, a novel algorithm is proposed to the detection of impulse noise point from images based on directional derivatives. First, the theory of calculus on directional derivatives is introduced in detail. Then it is applied to the field of image to removing noise with the discrete form derived from its continuous mathematical model. And a number of contrasting simulations illustrate that our algorithm not only can preserve the structure information while removing impulse noise but also can mostly save the gray value of the pixels undisturbed by noise. In addition, the comparisons of the filtering performance for removing impulse noise are analyzed in detail in the case of different noise densities, and also show that the algorithm suggested outperforms the conventional filter algorithms such as mean filter, median filter and so on in speed and impulse noise reduction, especially in random-valued impulse noise reduction. So it is a very good alternative to the existing schemes.

A Novel Fuzzy Direct Torque Control System for Three-level Inverter-fed Induction Machine

Diode clamped multi-level inverter （DCMLI） has a wide application prospect in high-voltage and adjustable speed drive systems due to its low stress on switching devices, low harmonic output, and simple structure. However, the problem of complexity of selecting vectors and capacitor voltage unbalance needs to be solved when the algorithm of direct torque control （DTC） is implemented on DCMLI. In this paper, a fuzzy DTC system of an induction machine fed by a three-level neutral-point-clamped （NPC） inverter is proposed. After introducing fuzzy logic, optimal selecting switching state is realized by applying various strategies which can distinguish the grade of the errors of stator flux linkage, torque, the neutral-point potential, and the position of stator flux linkage. Consequently, the neutral-point potential unbalance, the dr/dr of output voltage and the switching loss are restrained effectively, and desirable dynamic and steady-state performances of induction machines can be obtained for the DTC scheme. A design method of the fuzzy controller is introduced in detail, and the relevant simulation and experimental results have verified the feasibility of the proposed control algorithm.

Fundamental Concepts behind the Development of Gamma-Ray, Directed Energy Sources

We discuss novel advanced concepts suitable for the practical design of gamma-ray sources of directed energy. One concept is based on the self-channeling of a powerful optical laser in a gas within a metal tube. Another concept employs a direct excitation of a quadrupole nuclear level by a powerful optical laser. The third concept is based on the process of a high-order harmonic generation by an x-ray laser. All three concepts can be used for designing gamma-ray lasers that would have significant advantages over x-ray lasers. First, missile defense systems employing gamma-ray lasers would be weather independent. Second, the gamma-ray laser radiation can penetrate through the sand, which could be suspended in the air in a desert either naturally (due to strong winds) or artificially (as a protective “shield”). Besides, the first out of the three concepts can beemployed for creating non-laser gamma-ray sources of directed energy to be used for detecting stored radioactive materials, including the radioactive materials carried by an aircraft or a satellite. Last but not least: these concepts can be also used for remotely destroying biological and chemical weapons as a preemptive strike or during its delivery phase, as well as for distinguishing a nuclear warhead from decoy warheads. Thus, the defense capabilities of the proposed gamma-ray lasers can save numerous lives.

目标导向模式改善基层血液透析质量的临床研究

目的探讨目标导向模式在提升基层血液透析质量中的效果。方法以2021年10月—2022年3月于郑州大学第一附属医院南院区接受维持性血液透析治疗的基层患者为研究对象,观察目标导向模式管理6 m后,患者血管通路、临床指标、生活质量和卫生经济学指标等的改善情况。结果共纳入85例患者。与基线相比,6 m后带隧道带涤纶套导管的使用比例由42.35%降至27.06%(χ_(2)=6.303,P=0.043),单室尿素清除指数<1.2的患者比例降至7.05%(χ_(2)=21.182,P<0.001),透析间期体质量增加<5%干体质量的患者比例增至70.59%(χ_(2)=43.895,P<0.001),血红蛋白(χ_(2)=15.824,P=0.001)、血钾(χ_(2)=4.541,P=0.033)、血白蛋白(χ_(2)=31.224,P<0.001)、血磷(χ_(2)=17.306,P<0.001)及全段甲状旁腺激素(χ_(2)=17.162,P<0.001)达标率均提高;患者满意度由48.24%上升至100%(χ_(2)=59.365,P<0.001),生活质量评分提高(t=-22.455,P<0.001);直接医疗总成本无明显增加(t=1.387,P=0.167)。结论目标导向模式有效改善了基层患者的血液透析质量,利于患者并发症控制及生活质量提高,且不额外增加医疗花费,具有较强的应用价值。

高位定向钻孔抽采瓦斯搭接部位研究与应用

相邻钻场高位定向钻孔合理搭接是保证抽采瓦斯效果稳定、提高钻孔有效利用率的重要环节。根据顶板裂隙带高位定向钻孔抽采瓦斯机理研究结果,通过分析布置高位定向钻孔空间层位、煤层顶板裂隙带分布特征、钻孔轨迹等抽采效果影响因素,确定了实现成功接替的钻孔合理搭接部位,并在党家河煤矿进行了工程应用。应用显示:80~121 m钻孔搭接长度能够满足瓦斯抽采效果,在搭接部位外端点接替使每个钻孔减少钻孔工程量约40 m。

模块化多电平中压直流变换器的高可靠自供电架构研究

随着大容量中压直流配电技术的迅速发展,模块化多电平直流变换器受到工业界和学术界的广泛关注。然而,实现高可靠的辅助供电是实现直流变换器在中压直流配电场景中应用的关键技术瓶颈之一。为此,提出一种面向模块化多电平中压直流变换器的高可靠自供电系统架构。该架构实现主控制器、子模块等辅电系统的中压直联馈电,从而无需额外低压直流电源,并且大幅提升系统启动速度。基于此架构设计,模块化多电平直流变换器的子模块控制器和旁路开关辅助供电均实现3倍冗余,提高系统容错运行的能力。在子模块电容预充电阶段,提出一种基于辅助电源自主控制的子模块电容均压方法,保证模块化多电平直流变换器可靠预充电。最终通过实验验证所提辅助供电系统架构的可行性。

我国数字经济发展水平及区域网络结构

基于2013—2021全国30个省(区、市)的面板数据,利用CRITIC-熵权法测度省域数字经济发展水平,借助QAP分析法揭示各维度差异对数字经济发展水平差异的影响程度,并运用改进的修正引力模型,构建有向加权的省域数字经济关联网络,从整体、局部、个体3个维度量化其网络结构。研究发现,各省数字经济发展水平稳步提升,但不同省域间数字经济发展水平存在较大差异,北京、广东、上海数字经济发展水平遥遥领先;各省域数实融合平均水平变化最小,数字产业平均水平变化次之,数字基础设施和数字创新平均水平变化较大,数字基础设施差异和数字创新差异是数字经济发展水平差异的主要来源。从网络结构来看,各省份之间数字经济联系不够紧密,邻近省份更易形成凝聚子群,北京、上海、江苏、广东等省(市)占据网络中心位置,在数字经济关联网络中具有较强的影响力和吸引力。

考虑多重性能水准的特高压双柱悬索拉线塔风灾易损性分析

特高压双柱悬索拉线塔为典型的风敏感结构,其在不同强度风荷载作用下易遭受不同程度的损伤及破坏,针对此类结构的风致损伤评估研究亟待开展。该文提出一种考虑多重性能水准的特高压双柱悬索拉线塔风灾易损性分析方法。首先,开展结构立柱刚性模型的风洞试验,获取立柱的气动三分力系数并模拟得到相应风荷载;进而,定性划分双柱悬索拉线塔的性能水准,并基于Pushover分析给出各性能水准所对应的量化模型;最后,在求得双柱悬索拉线塔概率风灾需求模型的基础上,对其进行风灾易损性分析,并研究风向角对易损性曲线的影响。结果表明:结构主体立柱的气动升力系数CL和气动力矩系数CM在不同工况下均接近于0,气动阻力系数CD则均大于2,且从0°到45°呈递增趋势;基于立柱顶点位移可建立双柱悬索拉线塔的多重性能水准,而由风灾易损性分析可知,双柱悬索拉线塔在正常使用条件下发生各级风致损伤及破坏的概率很小,结构具有良好的抗风性能;风向角对双柱悬索拉线塔风灾易损性影响显著,结构的最不利和最有利风向角分别为90°和0°。

基于三矢量的PWM整流器改进模型预测直接功率控制方法

为解决传统模型预测直接功率控制存在的开关频率不恒定、计算量大、共模电压大问题,设计了一种三矢量模型预测直接功率控制方法。该方法通过在单个控制周期内输出三个不同电压矢量,克服了传统方法因连续多个周期输出相同电压矢量造成的开关频率不恒定问题;同时,在两相静止坐标系中,通过优化空间电压矢量图划分方法,采用两个代价函数确定电网电压矢量所在扇区,将算法的遍历次数从27次降至10次,克服了传统方法计算量大的问题;此外,通过去除两个冗余零矢量PPP与NNN,将共模电压抑制在Udc/3以内。最后搭建了5kW实验样机进行实验验证,仿真与实验结果表明,该方法可以有效降低计算量,输出谐波和共模电压抑制效果优于传统方法。

国产SW26010-Pro处理器上3级BLAS函数众核并行优化

BLAS(basic linear algebra subprograms)是最基本、最重要的底层数学库之一.在一个标准的BLAS库中,BLAS 3级函数涵盖的矩阵-矩阵运算尤为重要,在许多大规模科学与工程计算应用中被广泛调用.另外,BLAS 3级属于计算密集型函数,对充分发挥处理器的计算性能有至关重要的作用.针对国产SW26010-Pro处理器研究BLAS 3级函数的众核并行优化技术.具体而言,根据SW26010-Pro的存储层次结构,设计多级分块算法,挖掘矩阵运算的并行性.在此基础上,基于远程内存访问(remote memory access,RMA)机制设计数据共享策略,提高从核间的数据传输效率.进一步地,采用三缓冲、参数调优等方法对算法进行全面优化,隐藏直接内存访问(direct memory access,DMA)访存开销和RMA通信开销.此外,利用SW26010-Pro的两条硬件流水线和若干向量化计算/访存指令,还对BLAS 3级函数的矩阵-矩阵乘法、矩阵方程组求解、矩阵转置操作等若干运算进行手工汇编优化,提高了函数的浮点计算效率.实验结果显示,所提出的并行优化技术在SW26010-Pro处理器上为BLAS 3级函数带来了明显的性能提升,单核组BLAS 3级函数的浮点计算性能最高可达峰值性能的92%,多核组BLAS 3级函数的浮点计算性能最高可达峰值性能的88%.

月球直射辐照度仪同光轴装调及传递定标

采用激光水平仪与平面反射镜相结合的方式,在实验室模拟月球对追踪相机与探测镜头进行同光轴装调,避免未同光轴导致测试不到信号或者信号太小造成数据不准确的问题,实验结果显示实验室同光轴装调误差在±0.03°以内。为保证仪器定标系数的准确性,采用已溯源于国家计量基准的标准灯传递定标微光灯的方式,利用激光测距仪与激光水平仪、平面反射镜相结合的方式保证传递定标精度,解决了直接使用标准灯定标带来的探测器非线性响应误差和长定标距离误差的难题。为地基对月观测仪器建立高精度的月球辐射模型提供了一种思路与解决方法。

开绕组无刷双馈风力发电机直接功率控制容错策略

针对无刷双馈发电机提出了一种控制绕组采用开绕组拓扑(OW-BDFG)的馈电方案,将三相控制绕组两端打开分别连接变流器构成双两电平变流器(D-TLCs)馈电拓扑,对机侧变流器功率器件IGBT因故障切除后的容错控制策略进行深入分析,提出了功率偏差比较直接功率控制(PEC-DPC)的D-TLCs电压空间矢量优化方案。对比分析了8/4极25kW样机的仿真及实验结果,表明:当机侧变流器1个或2个IGBT被切除后,所提OW-BDFG的PEC-DPC仍可实现变速恒频、最大功率点跟踪及单位功率因数控制,进一步提高了该发电机系统的可靠性和鲁棒性。

空海跨域通信研究现状与发展趋势

伴随着人们对空中与海洋之间实时通信和数据传输需求的日益增长,研究稳健可靠的空海跨域通信技术非常重要。但是,由于空气与海水2种介质的不同特性以及恶劣环境的影响,使得空海跨域通信信道复杂多变,平稳地穿过空-水界面面临重大困难。为了更好地了解空海跨域通信技术发展脉络,文中对目前空海跨域通信研究进行了梳理与总结,全面阐述了其工作环境及通信技术发展情况,并将其分为空海跨域链路级直接通信和空海跨域中继通信两大类,而后分别从研究现状、面临困难、拟解决方法等方面对其进行详细总结及分析。最后,讨论了未来研究方向。文中所做研究可为带动空海跨域通信技术的发展提供助力。

英国A-level考试制度:实践模式、发展困境及改革路向

A-Level证书考试的实施是英国证书型大学入学考试制度确立的标志,既是英国教育的"黄金标准",亦是诸多英语授课大学招收新生的入学准则。英国自1951年实施A-level考试制度以来历经多次变革形成了较为完备的考试体系。然而,随着实践的深入,英国A-level入学考试逐步陷入应试的泥潭,严重影响其社会认可度和选拔功能的发挥,没能为中学生向大学过渡做好充足准备。因此,英国政府借改革契机调整考纲要求、转变评价机制、变更考试结构,为A-level入学考试制度发展寻找改良路径。英国A-level考试制度在转变培养模式、增强甄选效能、助力高大衔接方面所做的努力对我国高考制度改革的反观与思考不无裨益。

3D digital-image correlation insight into generalized relaxation behavior of sandstone under stress and pore pressure coupling

The occurrence of geological hazards and the instability of geotechnical engineering structures are closely related to the time-dependent behavior of rock.However,the idealization boundary condition for constant stress in creep or constant strain in relaxation is not usually attained in natural geological systems.Therefore,generalized relaxation tests that explore the simultaneous changes of stress and strain with time under different stress levels with constant pore-water pressure are conducted in this study.The results show that in area Ⅰ,area Ⅱ,and area Ⅲ,the stress and strain both change synchronously with time and show similar evolutionary laws as the strain-time curve for creep or the stress-time curve for relaxation.When the applied stress level surpasses the δ_(ci) or δ_(cd) threshold,the variations in stress and strain and their respective rates of change exhibit a significant increase.The radial deformation and its rate of change exhibit greater sensitivity in response to stress levels.The apparent strain deforms homogeneously at the primary stage,and subsequently,gradually localizes due to the microcrack development at the secondary stage.Ultimately,interconnection of the microcracks causes the formation of a shear-localization zone at the tertiary stage.The strain-time responses inside and outside the localization zone are characterized by local strain accumulation and inelastic unloading during the secondary and tertiary stages,respectively.The width of the shear-localization zone is found to range from 4.43 mm to 7.08 mm and increased with a longer time-to-failure.Scanning electron microscopy(SEM)reveals a dominant coalescence of intergranular cracks on the fracture surface,and the degree of physiochemical deterioration caused by water-rock interaction is more severe under a longer lifetime.The brittle sandstone’s time-dependent deformation is essentially controlled by microcrack development during generalized relaxation,and its expectancy-life is determined by its initial microstructural state and the rheological path.

多园区综合能源系统互联两阶段分布式运行优化策略

针对多个园区互联形成的多园区综合能源系统园区间电能交互、收益分配问题,首先,构建园区间电能交互运行框架;其次,从电力市场、隐私保护角度出发,提出一种两阶段分布式运行优化策略,第一阶段求解园区间的电能交互量,第二阶段基于非合作博弈理论出清园区互联运行模式下的电能交互成本与收益;最后,采用交替方向乘子法分离耦合变量依次对两阶段模型分布式求解。结果表明,交替方向乘子法在保护各主体隐私的前提下实现了两阶段模型的高效迭代求解,园区间通过电能交互整合资源降低了运行成本,经非合作博弈出清后的园区收益公平合理。

基于上下文方面记忆网络的方面级情感分析

基于深度记忆网络在方面级情感分析中的优势及针对相关工作中忽略上下文词与方面之间位置信息的不足,提出一种基于上下文方面记忆网络的方面级情感分析方法。不同记忆网络层关注上下文记忆不同部分,获取丰富方面感知上下文信息。为充分利用方面间信息,设计方面记忆网络更新模块,为所需方面生成邻近方面的语义和关系信息,在多计算层中的多头注意力机制输入采用两种策略计算上下文和方面词的相关性。在3个基准数据集上的实验结果表明,所提模型在性能评价指标的准确率和Macro-F1-score上与相关工作比较有一定提升。

计及多层级配电网的分布式新能源最大消纳空间分解测算

分布式新能源以“点多面广”的特征并入各级配电网,电网呈现新能源多层级接入、一体化消纳的特征。为促进新能源的充分消纳与高效利用,提出了一种多层级配电网新能源最大消纳空间测算模型,并将分布式新能源最大消纳空间测算问题转换为各层级配电网新能源最大消纳空间测算子问题,实现了各层级配电网分布式新能源最大消纳空间的精确测算。首先,以多层级配电网新能源接入量最大为目标函数,基于Distflow潮流模型建立多层级配电网分布式新能源消纳空间测算模型;然后,针对模型非凸以及求解效率低等问题,基于二阶锥松弛将模型转化为混合整数二阶锥规划模型,采用交替方向乘子法(alternating direction method of multipliers,ADMM),将多层级配电网新能源消纳空间测算问题转化为各级配电网新能源最大消纳空间子问题,将消纳空间模型转化为多层级配电网分布式新能源最大消纳空间分解测算模型;最后,以IEEE 6、7、9、10、12、15测试系统为例,验证该方法的有效性。