带有两种营销模式的供应链合作广告协调模型

本文考虑了带有直营店和代理商两种营销模式的Newsvendor型产品合作广告问题,分别建立了分散决策和集中控制下制造商和代理商的最优广告决策模型,分析了需求的随机性及直营店的营销模式对供应链双方广告决策的影响．给出了一个基于地方促销广告费用分摊的策略以实现供应链协调．最后,用数值仿真验证了本文的相关结论并得到一些重要的管理启示．     

The regulation of ion channels and transporters by glycolytically derived ATP

Glycolysis is an evolutionary conserved metabolic pathway that provides small amounts of energy in the form of ATP when compared to other pathways such as oxidative phosphorylation or fatty acid oxidation. The ATP levels inside metabolically active cells are not constant and the local ATP level will depend on the site of production as well as the respective rates of ATP production, diffusion and consumption. Membrane ion transporters (pumps, exchangers and channels) are located at sites distal to the major sources of ATP formation (the mitochondria). We review evidence that the glycolytic complex is associated with membranes; both at the plasmalemma and with membranes of the endo/sarcoplasmic reticular network. We examine the evidence for the concept that many of the ion transporters are regulated preferentially by the glycolytic process. These include the Na⁺/K⁺-ATPase, the H⁺-ATPase, various types of Ca²⁺-ATPases, the Na⁺/H⁺ exchanger, the ATP-sensitive K⁺ channel, cation channels, Na⁺ channels, Ca²⁺ channels and other channels involved in intracellular Ca²⁺ homeostasis. Regulation of these pumps, exchangers and ion channels by the glycolytic process has important consequences in a variety of physiological and pathophysiological processes, and a better understanding of this mode of regulation may have important consequences for developing future strategies in combating disease and developing novel therapeutic approaches. Received 20 July 2007; received after revision 30 July 2007; accepted 17 August 2007     

Conotoxins of the O-superfamily affecting voltage-gated sodium channels

The venoms of predatory cone snails harbor a rich repertoire of peptide toxins that are valuable research tools, but recently have also proven to be useful drugs. Among the conotoxins with several disulfide bridges, the O-superfamily toxins are characterized by a conserved cysteine knot pattern: C-C-CC-C-C. While ω-conotoxins and κ-conotoxins block Ca²⁺ and K⁺ channels, respectively, the closely related δ- and μO-conotoxins affect voltage-gated Na⁺ channels (Na_v channels). δ-conotoxins mainly remove the fast inactivation of Na_v channels and, thus, functionally resemble long-chain scorpion α-toxins. μO-conotoxins are functionally similar to μ-conotoxins, since they inhibit the ion flow through Na_v channels. Recent results from functional and structural assays have gained insight into the underlying molecular mechanisms. Both types of toxins are voltage-sensor toxins interfering with the voltage-sensor elements of Na_v channels. Received 27 December 2006; received after revision 30 January 2007; accepted 19 February 2007     

革命历史题材剧对相关红色旅游景点网络关注度的影响研究——以《觉醒年代》为例

革命历史题材剧在促进大众情感认知和文化认同、激发旅游需求、增强红色旅游景点吸引力等方面发挥着重要作用.该文以重大革命历史题材剧《觉醒年代》为例,以2020—2021年相关红色旅游景点百度指数为数据基础,采用赫芬达尔指数、空间自相关、回归分析等方法探讨革命历史题材剧对其相关红色景点网络关注度时空特征的影响.研究发现：1)《觉醒时代》播出后,相关红色景点网络关注度与播出前同期水平相比显著增加,且非知名景点的增长幅度大于知名景点;年内均衡程度变动较小,年内分布趋势呈现倒“U”型分布;2)《觉醒时代》播出后,相关红色景点网络关注度空间集聚程度提升,冷热点空间格局较为稳定,呈现胡焕庸线分割特征,热点区主要集聚在长江中游城市群及邻近区域;延乔路、龙华烈士陵园、毛泽东故居网络关注度的空间响应较为明显;长江中游城市群网络关注度的提升较为显著;3)《觉醒年代》相关联旅游景点的平均每日互联网关注度与电视剧的受欢迎程度呈正相关.     

基于深度学习的MIMO系统联合信号检测与信道译码

为进一步提高多输入多输出(multiple-input multiple-output,MIMO)系统性能,研究了深度学习方法来联合解决MIMO系统信号检测与信道译码问题.通过将深度神经网络、自动编码器神经网络与传统MIMO通信系统的物理层架构进行有机融合,构建了基于神经网络的MIMO系统模型,可获取系统发射端的信息比...     

基于注意力机制的分层次交互融合多模态情感分析

针对基于视频的多模态情感分析中,通常在同一语义层次采用同一种注意力机制进行特征捕捉,而未能考虑模态间交互融合对情感分类的差异性,从而导致模态间融合特征提取不充分的问题,提出一种基于注意力机制的分层次交互融合多模态情感分析模型(hierarchical interactive fusion network based on attention mechanism, HFN-AM),采用双向门控循环单元捕获各模态内部的时间序列信息,使用基于门控的注意力机制和改进的自注意机制交互融合策略分别提取属于句子级和篇章级层次的不同特征,并进一步通过自适应权重分配模块判定各模态的情感贡献度,通过全连接层和Softmax层获得最终分类结果。在公开的CMU-MOSI和CMU-MOSEI数据集上的实验结果表明,所给出的分析模型在2个数据集上有效改善了情感分类的准确率和F1值。     

可重构智能表面辅助大规模MIMO通信波束成形设计

可重构智能表面(reconfigurable intelligent surface, RIS)在无线通信系统中能够提高信道性能。为了降低传统波束成形方案中功率分配的复杂度,结合毫米波(millimeter wave, mmWave)特点建立信道模型,提出一种基于正交匹配追踪(orthogonal matching pursuit, OMP)和统一信道分解(uniform channel decomposition, UCD)的RIS辅助mmWave多输入多输出(multiple input multiple output, MIMO)系统混合波束成形设计方案。利用视距(line of sight, LoS)信道的到达角(angle of arrival, AoA)和离开角(angle of departure, AoD)设计了RIS的反射矩阵。该方法不需要复杂的功率分配,能够使得多个并行数据流具有相同的信道增益。仿真结果表明,该混合波束成形方案具有更好的误比特率(bit error rate, BER)性能及频谱效率。     

基于改进蚁群算法和共用航段的搜索救援通道规划

针对战斗搜救任务场景中,空中搜索救援等待区至跳伞区之间的往返路径规划问题,提出一种基于改进蚁群算法和共用航段的搜索救援通道规划方法。首先,针对基本蚁群算法易出现局部极值及收敛速度慢等缺点,对启发函数、状态转移策略、信息素更新策略进行了改进。然后,针对单起点多目标点的路径规划问题,提出了设置分航点以生成共用航段的搜索救援通道规划方法,并以总路径长度表征任务准备负荷量与分别规划航路进行对比。实验结果表明,在本文测试环境下,改进蚁群算法规划的路径长度分别缩减15.4%和14.2%,搜索救援通道的任务准备负荷量减少25.4%。所提出的搜索救援通道规划方法有一定的理论和应用价值。     

基于YOLOv5s-FCS的钢材表面缺陷检测

针对传统钢材表面缺陷检测方法易出现误检、漏检和部分缺陷种类检测精度低等问题,本文设计了一种钢材表面缺陷网络YOLOv5s-FCS。首先本文引用了FReLU激活函数构建了卷积模块CBF,有效增强了网络的空间解析能力,优化了网络检测精度;其次,本文将坐标注意力机制嵌入到网络的neck部分来增强网络特征融合的能力,从而使网络能够提取更加丰富的特征信息;最后,将YOLOv5s的损失函数替换为SIoU loss,提高了预测框的回归精度。通过在NEU-DET数据集上进行消融实验、可视化对比实验,结果表明,YOLOv5s-FCS网络的mAP值达到了0.747,相较于原YOLOv5s网络提高了8.3%,相较于YOLOv3网络提高了11.8%,相较于YOLOXs网络提高了4.2%,相较于YOLOv6s提高了1.4%,验证了该方法的可行性、有效性。     

基于信道状态信息反馈的特征波束形成算法

在多输入多输出通信系统中,为了有效地降低系统的误符号率,提出了一种特征波束形成算法。在假设接收端能准确地完成信道估计,并能及时将信道的互相关矩阵反馈到发射端的基础上,研究了接收和发射均采用多天线的特征波束形成算法,对其性能进行了详细的分析和推导,并在此基础上研究了波束形成与空时块码的结合以提高系统的频谱利用率。计算机仿真结果表明,在接收端采用多根天线相对于单根天线能明显降低系统的误符号率。特征波束形成算法通过同时提供波束形成增益和分集增益位置,有效地提高了系统性能。