Ripple-effect analysis for operational architecture of air defense systems with supernetwork modeling

In order to solve the problem that the ripple-effect analy- sis for the operational architecture of air defense systems （OAADS） is hardly described in quantity with previous modeling approaches, a supernetwork modeling approach for the OAADS is put for- ward by extending granular computing. Based on that operational units and links are equal to different information granularities, the supernetwork framework of the OAADS is constructed as a “four- network within two-layer” structure by forming dynamic operating coalitions, and measuring indexes of the ripple-effect analysis for the OAADS are given combining with Laplace spectral radius. In this framework, via analyzing multidimensional attributes which inherit relations between operational units in different granular scales, an extended granular computing is put forward integrating with a topological structure. Then the operation process within the supernetwork framework, including transformation relations be- tween two layers in the vertical view and mapping relations among functional networks in the horizontal view, is studied in quantity. As the application case shows, comparing with previous modeling approaches, the supernetwork model can validate and analyze the operation mechanism in the air defense architecture, and the ripple-effect analysis can be used to confirm the key operational unit with micro and macro viewpoints.

Radial Basis Function Neural Network Based Super- Resolution Restoration for an Undersampled Image

To achieve restoration of high frequency information for an undersampled and degraded low-resolution image, a nonlinear and real-time processing method-the radial basis function (RBF) neural network based super-resolution method of restoration is proposed. The RBF network configuration and processing method is suitable for a high resolution restoration from an undersampled low-resolution image. The soft-competition learning scheme based on the k-means algorithm is used, and can achieve higher mapping approximation accuracy without increase in the network size. Experiments showed that the proposed algorithm can achieve a super-resolution restored image from an undersampled and degraded low-resolution image, and requires a shorter training time when compared with the multiplayer perception (MLP) network.

Super Node Routing Strategy in Content-Centric Networking

There were two strategies for the data forwarding in the content-centric networking(CCN): forwarding strategy and routing strategy. Forwarding strategy only considered a separated node rather than the whole network performance, and Interest flooding led to the network overhead and redundancy as well. As for routing strategy in CCN, each node was required to run the protocol. It was a waste of routing cost and unfit for large-scale deployment.This paper presents the super node routing strategy in CCN. Some super nodes selected from the peer nodes in CCN were used to receive the routing information from their slave nodes and compute the face-to-path to establish forwarding information base(FIB). Then FIB was sent to slave nodes to control and manage the slave nodes. The theoretical analysis showed that the super node routing strategy possessed robustness and scalability, achieved load balancing,reduced the redundancy and improved the network performance. In three topologies, three experiments were carried out to test the super node routing strategy. Network performance results showed that the proposed strategy had a shorter delay, lower CPU utilization and less redundancy compared with CCN.

Super-resolution image reconstruction based on three-step-training neural networks

A new method of super-resolution image reconstruction is proposed, which uses a three-step-training error backpropagation neural network （BPNN） to realize the super-resolution reconstruction （SRR） of satellite image. The method is based on BPNN. First, three groups learning samples with different resolutions are obtained according to image observation model, and then vector mappings are respectively used to those three group learning samples to speed up the convergence of BPNN, at last, three times consecutive training are carried on the BPNN. Training samples used in each step are of higher resolution than those used in the previous steps, so the increasing weights store a great amount of information for SRR, and network performance and generalization ability are improved greatly. Simulation and generalization tests are carried on the well-trained three-step-training NN respectively, and the reconstruction results with higher resolution images verify the effectiveness and validity of this method.

学习驱动的图像压缩算法研究

目前,基于卷积神经网络(CNN)深度学习的图像压缩已经取得了优异的成果,但是CNN的感受野通常受限,不能学习到图像非局部区域内像素之间的上下文关系,缺少了长距离建模和感知能力,容易造成结果失真、伪影和压缩率较高等问题。针对以上问题,提出2种解决办法:设计一种由CNN、多尺度注意力(MSA)机制和残差单元构成的对称编解码器架构,该架构在对图片进行通道变换和空间变换的同时引入多尺度注意力机制,能够对特征进行重新校准,减少潜在表示的冗余像素;设计一种基于U型框架的超先验网络,可以在不同层级上获取多尺度的上下文信息,在帮助提取高级语义特征的同时,保留详细的低级特征信息,能够更好地进行边界细化和细节恢复。在Kodak、Tecnick和CLIC这3种数据集上将所提方法与其他先进的图像压缩方法进行对比实验,结果表明,在相同比特率下,该方法相较对比方法在峰值信噪比(PSNR)指标上分别提高了约0.3 dB、0.6 dB、0.5 dB。所提方法在保证压缩率的同时,能够有效提高对非重复纹理特征和图像细节特征的重建效果。

Bootstrapped Multi-Model Neural-Network Super-Ensembles for Wind Speed and Power Forecasting

The bootstrap resampling method is applied to an ensemble artificial neural network (ANN) approach (which combines machine learning with physical data obtained from a numerical weather prediction model) to provide a multi-ANN model super-ensemble for application to multi-step-ahead forecasting of wind speed and of the associated power generated from a wind turbine. A statistical combination of the individual forecasts from the various ANNs of the super-ensemble is used to construct the best deterministic forecast, as well as the prediction uncertainty interval associated with this forecast. The bootstrapped neural-network methodology is validated using measured wind speed and power data acquired from a wind turbine in an operational wind farm located in northern China.

基于ViT-CNN特征增强的图像超分辨率

卷积神经网络(CNN)可以提取图像的局部相关特征,视觉Transformer(ViT)则侧重于捕获图像的远距离依赖关系,二者有效结合能够改进图像的重构质量.研究了一种基于ViT-CNN特征增强的图像超分辨率(SR)网络.具体来说,网络包含了基于ViT的SR分支与基于CNN的梯度分支,SR分支主要用于提取图像特征域中的全局相关性,而梯度分支则专注于图像梯度域中的局部依赖关系.通过对两种信息的融合与渐进增强,获得高倍放大的重构图像.此外,在网络的学习阶段引入了梯度损失及渐进训练策略,有效降低了网络的训练难度并增强了训练的稳定性.在多个公开数据集上的大量实验结果验证了所提方法在改善重构系统性能方面的有效性.

基于Super SAB神经网络算法的主变压器故障诊断模型

人工神经网络(ANN)由于其高度的非线性映射能力在电力系统模式识别及非线性优化领域有着广泛深入的应用研究。本文将Super SAB神经网络算法应用于主变压器溶解气体故障诊断(DGA),通过与带动量因子的标准BP算法、Bold Driver算法比较,验证Super SAB算法在故障模式识别中具有更好的学习效率与泛化能力,故障诊断的准确度高于传统分析方法,表明其在变电设备状态诊断中具有良好的应用前景。

基于深度学习的复合超分辨率重建算法在膝关节MRI中的临床应用价值

目的:探讨临床环境中通过优化扫描参数结合基于深度学习的复合超分辨率重建算法在提升膝关节MRI扫描效率和图像质量的可行性。方法:前瞻性搜集110例行膝关节MRI平扫的患者,先后进行常规(常规组)与复合超分辨率重建算法扫描(复合组),采用双盲法比较两组主客观图像质量。结果:相较常规组,复合组PD和T1序列的骨髓、软骨、半月板、韧带、肌肉、脂肪、关节液的SNR分别提升89.3%、52.5%、65.3%、73.8%、60.3%、103.9%、58.9%和78.0%、172.9%、78.0%、72.5%、75.4%、63.4%、97.0%。相较常规组,复合组PD和T1序列的软骨-关节液、软骨-骨髓、半月板-关节液、韧带-关节液、骨髓-关节液、脂肪-关节液、肌肉-关节液的CNR分别提升119.5%、83.3%、116.2%、109.2%、109.2%、99.3%、116.8%和61.7%、23.1%、78.7%、32.5%、161.7%、44.9%、39.2%。复合组的峰值信噪比(PSNR)相较常规组显著提高(P<0.05),结构相似度(SSIM)均>0.999。主观图像质量评价中复合组病灶边缘区分度、运动伪影和综合诊断度的主观评分显著高于常规组(P<0.05),两组病灶辨别度的主观评分差异无统计学意义(P>0.05)。结论:合理优化扫描参数并结合基于深度学习的复合超分辨率重建算法可在提升扫描效率的同时显著提高膝关节MRI的图像质量和综合诊断效果。

基于ANP和Super Decision的通信链路威胁评估排序

运用ANP（网络层次分析）方法对通信链路威胁进行评估排序,建立了ANP控制层和网络层的模型,并使用Super Decision计算极限超矩阵,得到最终目标排序。由于考虑到了评估指标的相互影响关系,ANP法比传统AHP方法更为接近实际情况。

基于U模型的非线性系统Super-Twisting滑模控制研究

为了对基于U模型的非线性控制系统进行研究,利用Super-Twisting控制算法,解决非仿射非线性系统的控制问题,对非线性函数进行神经网络逼近,运用Super-Twisting控制算法进行控制。选取恰当的Lyapunov函数,对Super-Twisting算法的收敛性进行了证明。为了验证该方法的可行性和有效性,利用Matlab软件进行仿真,结果表明在神经网络自适应Super-Twisting控制器的作用下,被控系统具有快速的跟踪性能和输出的有界性。

基于Super-node结构的P2P网络的信誉评价系统

针对基于Super-node结构的P2P网络提出了一个新的简单的信誉评价系统。该系统除了能根据其它对等点的反馈信息来建立对等点之间的信任关系外,还能帮助识别一些恶意对等点,从而保证整个P2P网络的正常运行。

面向城乡客货一体化运输的多车型公交混合调度研究

针对农村地区乘客出行需求密度低,客运公交无法充分利用其运力,在平峰期甚至出现空载现象,将客运和货运结合,以不同的客货比例形成多车型公交,能有效利用空闲运力。本文基于超级时空网络提出多车型公交混合调度模型,并设计了3M(Mixed creation, Mutation and Mature operators)进化算法。以南里奥格兰德州阿雷格里港运输数据进行实例研究,同时进行敏感性分析,探究不同车型、折旧成本系数和种群数量对总成本与车队规模的影响。结果表明,当车型偏向以乘客为主或以货物为主时,总成本均保持在较低的水平;进一步分析发现车型偏向以货物为主时,整体车队数量减少。客运和货运相结合模式,不仅有效提升了车辆利用率,同时也降低近21.93%的运营成本。

自由贸易试验区建设能提升城市绿色创新效率吗?——基于创新价值链视角的经验证据

自由贸易试验区建设以绿色为底色,以创新为驱动,是促进经济高质量发展的催化剂。选取2006—2022年中国284个地级市的面板数据,基于创新价值链视角研究自由贸易试验区建设对城市绿色创新效率以及子阶段效率的直接影响、影响机制和影响的空间溢出效应,得出以下研究结论:(1)自由贸易试验区建设能够提高城市绿色创新效率以及绿色科技研发和绿色成果转化两个子阶段效率,这一结论经过双重机器学习再检验和内生性分析等一系列检验依旧稳健;(2)这一高水平制度性开放举措在沿海城市、高营商环境城市和低环境规制城市的政策效应更显著;(3)提高城市创业活跃度、促进创新人才流动和加强战略新兴产业集聚是自由贸易试验区建设提高城市绿色创新效率以及子阶段效率的作用机制;(4)自由贸易试验区建设对周围城市绿色创新效率和绿色科技研发效率产生显著的正向影响,且正向空间溢出的衰减边界分别为350千米和200千米,而自由贸易试验区建设对周围城市绿色成果转化效率的影响不显著。在此基础上,为中国推进自由贸易试验区战略分层次、多维度和分阶段地提升城市绿色创新效率提供政策建议。

802.11 Mesh network中WBDMR分布式路由协议的研究

提出了一种适用于中等规模的802.11Meshnetwork网络的无线分布式路由协议(WBDMR)。该协议借鉴了基于Ad-Hoc的ADOV协议并进行了改造,实现了不同BSS(基本服务集)间无线客户端通过无线中继相互通信。利用节点间交换链路损耗信息实现了动态负载平衡、有效地抑制了冗余路由信息的传播、消除了路由环路、建立了备份路由。仿真结果表明,WBDMR协议具有较好的路由特性和时延特性。

基于超分辨率重建算法的二氧化碳浓度场研究

针对目前激光吸收光谱技术在非均匀分布烟道条件下获取CO_(2)初始信号进行浓度场重建时存在的计算时间长、重建精度低等问题,提出了一种基于超分辨率算法的CO_(2)浓度场重建方法,并对EO算法进行了优化改进。改进后的EO算法在CO_(2)浓度分布的计算中显著提升了迭代速度和搜寻精度;在超分辨率温度场重建算法方面,加入了输入补偿机制并优化了密集连接网络。经过实验对比验证,改进的CO_(2)浓度重建算法有效提高了重建精度,平均ARE与RMSE值分别达到了0.053733和0.054。

Super VLAN技术在NGN承载网络优化中的应用

本文分析了传统VLAN技术和Super VLAN技术在网络规划及IP地址分配上的优劣。结合目前AG承载网络的组网模式存在着网络风暴的安全问题及IP地址浪费的情况,文中重点讲述了基于Super VLAN技术进行网络优化的方法,为NGN的承载网络安全及IP地址的配置提供了可靠的解决方案。

基于两阶段Super-NSBM模型的城市生态福利绩效评价研究

如何在生态门槛内为城市居民实现福利水平的持续提升即城市生态福利绩效的提高是城市生态文明建设和可持续发展的根本要求和必然选择。本文从生态福利视角出发,基于DEA方法构建城市生态福利绩效评价指标体系,并运用两阶段考虑非合意产出的Super-NSBM模型和DEA窗口分析法测算了中国35个主要城市2011—2015年的生态福利绩效,将生态福利绩效分解为生态经济效率和经济福利效率两阶段,打开城市生态福利转化过程的“黑箱”,并通过冗余度分析为各城市指明效率改进的方向。研究发现:“十二五”期间,我国35个主要城市的整体生态福利绩效水平不高,且城市之间差异较大,总体上未能实现DEA有效,但近几年呈现出逐年改善的利好趋势。分阶段效率来看,第二阶段的经济福利效率高于第一阶段的生态经济效率,生态经济效率偏低是导致整体生态福利绩效水平不高的主要原因。分区域来看,城市生态福利绩效综合水平呈现出“中部最高、东部次之、西部最低”的态势。与传统的单阶段(“黑箱”评价)视角下考虑非合意产出Super-SBM模型相比,两阶段视角下基于考虑非合意产出的超效率网络SBM模型的测算结果精确度更高,再次验证了本文选取该模型的合理性。为提高城市生态福利绩效,促进城市生态文明建设与可持续发展,提出低效率城市应加快产业结构转型、发展绿色生态产业;有效推进京津冀城市群协同发展;建设公共交通导向的紧凑型低碳城市等建议。

RIS辅助的OFDM系统中时变信道估计方法

为了克服在可重构智能反射面(reconfigurable intelligent surface,RIS)辅助的正交频分复用(orthogonal frequency division multiplexing,OFDM)系统中现有基于深度学习的信道估计方法计算复杂度过高的问题,在RIS下利用基扩展模型(base extension model,BEM)对时变信道进行建模,并提出基于残差链接超分辨率卷积神经网络的时变信道估计方法。具体来说,所提方法首先将参数较多的信道系数估计转换为参数较少的基系数估计,以降低所提方法计算复杂度。在线下训练中,利用低分辨率的基系数估计对神经网络进行训练,仅需要少量的输入即可获取高分辨率的信道估计。为了提高所提方法的实用性,将网络训练的标签设置为具有高精度的信道估计值,而非理想的信道信息。仿真实验验证,所提方法在RIS辅助移动通信系统下能够准确获取时变信道信息,且具有更高的估计精度和更低的计算复杂度。

The New Architecture with Time-Spatial Consistency for 5G Networks

While operators have started deploying fourth generation(4G) wireless communication systems,which could provide up to1 Gbps downlink peak data rate,the improved system capacity is still insufficient to meet the drastically increasing demand of mobile users over the next decade.The main causes of the above-mentioned phenomenon include the following two aspects:1) the growth rate of the network capacity is far below that of user's demand,and 2) the relatively deterministic wireless access network(WAN) architecture in the existing systems cannot accommodate the prominent increase of mobile traffic with space-time domain dynamics.In order to address the above-mentioned challenges,we investigate the time-spatial consistency architecture for the future WAN,whilst emphasizing the critical roles of some spectral-efficient techniques such as Massive multiple-input multiple-output(MIMO),full-duplex(FD)operation and heterogeneous networks(HetNets).Furthermore,the energy efficiency(EE)of the HetNets under the proposed architecture is also evaluated,showing that the proposed user-selected uplink power control algorithm outperforms the traditional stochastic-scheduling strategy in terms of both capacity and EE in a two-tier HetNet.The other critical issues,including the tidal effect,the temporal failure owing to the instantaneously increased traffic,and the network wide load-balancing problem,etc.,are also anticipated to be addressed in the proposed architecture.(Abstract)