Totally Coded Method for Signal Flow Graph Algorithm

After a code-table has been established by means of node association information from signal flow graph, the totally coded method (TCM) is applied merely in the domain of code operation beyond any figure-earching algorithm. The code-series (CS) have the holo-information nature, so that both the content and the sign of each gain-term can be determined via the coded method. The principle of this method is simple and it is suited for computer programming. The capability of the computer-aided analysis for switched current network (SIN) can be enhanced.

RECONSTRUCTION OF ONE DIMENSIONAL MULTI-LAYERED MEDIA BY USING A TIME DOMAIN SIGNAL FLOW GRAPH TECHNIQUE

A novel inverse scattering method to reconstruct the permittivity profile of one-dimensional multi-layered media is proposed in this paper.Based on the equivalent network ofthe medium,a concept of time domain signal flow graph and its basic principles are introduced,from which the reflection coefficient of the medium in time domain can be shown to be a series ofDirac δ-functions(pulse responses).In terms of the pulse responses,we will reconstruct both thepermittivity and the thickness of each layer will accurately be reconstructed.Numerical examplesverify the applicability of this

Development of an Improved GUI Automation Test System Based on Event-Flow Graph

A more automated graphic user interface (GUI) test model, which is based on the event-flow graph, is proposed. In the model, a user interface automation API tool is first used to carry out reverse engineering for a GUI test sample so as to obtain the event-flow graph. Then two approaches are adopted to create GUI test sample cases. That is to say, an improved ant colony optimization (ACO) algorithm is employed to establish a sequence of testing cases in the course of the daily smoke test. The sequence goes through all object event points in the event-flow graph. On the other hand, the spanning tree obtained by deep breadth-first search (BFS) approach is utilized to obtain the testing cases from goal point to outset point in the course of the deep regression test. Finally, these cases are applied to test the new GUI. Moreover, according to the above-mentioned model, a corresponding prototype system based on Microsoft UI automation framework is developed, thus giving a more effective way to improve the GUI automation test in Windows OS.

Analysis of Electronic Circuits with the Signal Flow Graph Method

In this work a method called “signal flow graph (SFG)” is presented. A signal-flow graph describes a system by its signal flow by directed and weighted graph;the signals are applied to nodes and functions on edges. The edges of the signal flow graph are small processing units, through which the incoming signals are processed in a certain form. In this case, the result is sent to the outgoing node. The SFG allows a good visual inspection into complex feedback problems. Furthermore such a presentation allows for a clear and unambiguous description of a generating system, for example, a netview. A Signal Flow Graph (SFG) allows a fast and practical network analysis based on a clear data presentation in graphic format of the mathematical linear equations of the circuit. During creation of a SFG the Direct Current-Case (DC-Case) was observed since the correct current and voltage directions was drawn from zero frequency. In addition, the mathematical axioms, which are based on field algebra, are declared. In this work we show you in addition: How we check our SFG whether it is a consistent system or not. A signal flow graph can be verified by generating the identity of the signal flow graph itself, illustrated by the inverse signal flow graph (SFG−1). Two signal flow graphs are always generated from one circuit, so that the signal flow diagram already presented in previous sections corresponds to only half of the solution. The other half of the solution is the so-called identity, which represents the (SFG−1). If these two graphs are superposed with one another, so called 1-edges are created at the node points. In Boolean algebra, these 1-edges are given the value 1, whereas this value can be identified with a zero in the field algebra.

A Survey of Malware Classification Methods Based on Data Flow Graph

Malware is emerging day by day.To evade detection,many malware obfuscation techniques have emerged.Dynamicmalware detectionmethods based on data flow graphs have attracted much attention since they can deal with the obfuscation problem to a certain extent.Many malware classification methods based on data flow graphs have been proposed.Some of them are based on userdefined features or graph similarity of data flow graphs.Graph neural networks have also recently been used to implement malware classification recently.This paper provides an overview of current data flow graph-based malware classification methods.Their respective advantages and disadvantages are summarized as well.In addition,the future trend of the data flow graph-based malware classification method is analyzed,which is of great significance for promoting the development of malware detection technology.

A Method for Software Vulnerability Detection Based on Improved Control Flow Graph

With the rapid development of software technology, software vulnerability has become a major threat to computer security. The timely detection and repair of potential vulnerabilities in software, are of great significance in reducing system crashes and maintaining system security and integrity. This paper focuses on detecting three common types of vulnerabilities: Unused_Variable, Use_of_Uninitialized_Variable, and Use_After_ Free. We propose a method for software vulnerability detection based on an improved control flow graph(ICFG) and several predicates of vulnerability properties for each type of vulnerability. We also define a set of grammar rules for analyzing and deriving the three mentioned types of vulnerabilities, and design three vulnerability detection algorithms to guide the process of vulnerability detection. In addition, we conduct cases studies of the three mentioned types of vulnerabilities with real vulnerability program segments from Common Weakness Enumeration(CWE). The results of the studies show that the proposed method can detect the vulnerability in the tested program segments. Finally, we conduct manual analysis and experiments on detecting the three types of vulnerability program segments(30 examples for each type) from CWE, to compare the vulnerability detection effectiveness of the proposed method with that of the existing detection tool Cpp Check. The results show that the proposed method performs better. In summary, the method proposed in this paper has certain feasibility and effectiveness in detecting the three mentioned types of vulnerabilities, and it will also have guiding significance for the detection of other common vulnerabilities.

Cyclic Reconfigurable Flow Shop under Different Configurations Modeling and Optimization Based on Timed Event Graph

Based on the idea that modules are independent of machines, different combinations of modules and machines result in different configurations and the system performances differ under different configurations, a kind of cyclic reconfigurable flow shops are proposed for the new manufacturing paradigm-reconfigurable manufacturing system. The cyclic reconfigurable flow shop is modeled as a timed event graph. The optimal configuration is defined as the one under which the cyclic reconfigurable flow shop functions with the minimum cycle time and the minimum number of pallets. The optimal configuration, the minimum cycle time and the minimum number of pallets can be obtained in two steps.

面向交通流量预测的时空Graph-CoordAttention网络

交通预测是城市智能交通系统的一个重要研究组成部分,使人们的出行更加效率和安全。由于复杂的时间和空间依赖性,准确预测交通流量仍然是一个巨大的挑战。近年来,图卷积网络(GCN)在交通预测方面表现出巨大的潜力,但基于GCN的模型往往侧重于单独捕捉时间和空间的依赖性,忽视了时间和空间依赖性之间的动态关联性,不能很好地融合它们。此外,以前的方法使用现实世界的静态交通网络来构建空间邻接矩阵,这可能忽略了动态的空间依赖性。为了克服这些局限性,并提高模型的性能,提出了一种新颖的时空Graph-CoordAttention网络(STGCA)。具体来说,提出了时空同步模块,用来建模不同时刻的时空依赖交融关系。然后,提出了一种动态图学习的方案,基于车流量之间数据关联,挖掘出潜在的图信息。在4个公开的数据集上和现有基线模型进行对比实验,STGCA表现了优异的性能。

基于循环独立机制的交通流量预测

交通流量预测是智能交通控制和管理系统的一个重要环节,但交通流量数据具有时间和空间上的非线性和复杂性等特征,为对其进行精准预测,本文提出了Graph Temopral Recurrent Independent Mechanisms (G-tRIM)模型。该模型使用图注意力网络(Graph Attention Networks, GAT)来有效捕获交通流量数据的空间依赖关系,使用循环独立机制(Recurrent Independent Mechanisms, RIM)来精准刻画交通流量数据的潜在状态。最后在北京和贵州数据集上,以均方误差(Mean Square Error, MSE)和平均绝对误差(Mean Absolute Error, MAE)为指标进行实验,结果表明,G-tRIM在各个数据集上的表现均优于基准模型。

多尺度融合与动态自适应图的公交客流预测模型

公交客流预测是公共交通规划和管理中的重要问题。虽然时空图卷积在地铁客流预测任务中获得了很好的预测效果,但是面对公交更复杂的线路、大规模的节点数据,现有的基于图卷积的空间建模方法将带来巨大的空间内存消耗。同时,公交客流量短时间范围内更可能受到瞬时交通状况的影响。为了解决这些挑战,提出了一种多尺度融合和动态自适应图的公交客流预测模型(MFDAG)。该模型融合客流、时刻和周信息以增加数据的特征维度,用动态自适应图的方法来学习不同站点之间的关系。进一步提出了一种多尺度融合传播的方法来表示复杂的空间依赖关系,同时设计了一种多尺度卷积传播的方法来学习不同尺度的时间依赖关系。在两个真实的客流数据集上进行了实验,并与其他交通预测方法进行了比较。实验结果表明,所提出的多尺度融合和动态自适应图的公交客流预测方法具有更高的预测准确度。

基于改进Node2vec算法的锅炉温度场分割方法研究

针对温度场特征参数差异引发的锅炉温度场分割准确性的问题,以维持温度场特征为目标,引入图结构表达场数据,通过改进Node2vec算法进行聚类分析,进而实现锅炉温度场的最佳分割。该方法基于多维度的特征信息对锅炉温度场实现分割,能够更准确地保留流场特征。在标准数据集上进行了实验验证,结果表明在具有多维度特征的数据集上,所提方法相比其他对比算法在分割效果方面有提升显著。最后将提出的方法用于分割电站锅炉温度场,结果表明该方法可以很好地捕捉温度场数据中的局部和全局特征,且结果具有较好的精确性。

基于异构指令图的恶意软件分类方法研究

恶意软件是当今互联网上最大的安全威胁之一。目前基于图深度学习的恶意软件分类研究未考虑同家族恶意软件的控制流信息所隐含的内在相似性。针对该问题提出了一种基于异构指令图的恶意软件分类方法MCHIG,包括三个阶段:异构指令图生成、结点嵌入和恶意软件分类,首先生成数据集MyHIG,再应用GraphSAGE对不同类型的边分别进行消息传递,完成异构指令图文件结点分类和指令结点嵌入,最后基于控制流图完成恶意软件分类任务。在BIG2015数据集上嵌入阶段的分类精度达到97.81%,分类阶段分别进行了五折和十折交叉验证,其中十折交叉验证的性能更佳,准确度达到99.91%,在BODMAS_mini少样本数据集上,在嵌入阶段达到96.53%,在分类阶段达到98.76%,优于目前较先进的其他恶意软件分类模型。

融合多种模态特征的井下供水管网流量预测

煤矿井下供水系统是煤矿安全生产的生命线,供水管网水流量的预测是供水系统优化调度的基础,预测的重要性对供水调度有重要影响。文章提出了一种融合多模态数据特征的煤矿井下供水管网流量预测方法,该方法通过图深度学习的方法实现了对井下管网空间拓扑结构、历史时间依赖、井下实际生产工况、周期相关等多种数据模态特征的融合,具体的,使用添加空间注意机制的图卷积神经网络获取井下管网监测点的空间拓扑关系,然后利用循环神经网络中的门控循环单元获取监测点的时间依赖,并融合煤矿生产规律与不同周期的流量数据形成最终预测结果,通过陕西亭南煤矿实际数据进行实验,结果表明,提出的预测方法相较于SVM、LSTM、STGCN等方法能更准确地预测井下流量未来的趋势,预测偏差分别降低了9.3%、6.84%和3.65%。

基于图对比学习的稳健交通流量预测

作为智能交通系统的核心技术,稳健的交通流量预测是一个长期存在但具有挑战的任务。不论是工业界还是学术界,现有模型需要大量训练数据、易受数据噪声影响而表现不稳健,是限制该领域发展与实际应用落地的重要因素。而在学术界,图对比学习可以通过数据增强与对比损失来降低数据需求量,同时提升模型抵抗数据噪声的能力。提出一种交通流量图对比学习(TFGCL)框架,用于稳健的交通流量预测。TFGCL框架有3个创新点:针对交通流量图(TFG)数据的独特时空特性,TFGCL框架从时间和空间2个角度出发,提出3种TFG数据增强方法。针对TFG数据中语义相似的假负样本,提出一个过滤策略使TFGCL框架能够免受其干扰,从而学习到高质量的表征。TFGCL框架通过联合交通流量预测任务和图对比学习任务进行同时训练。在3个真实交通数据集上与8个基线模型进行对比实验,结果表明:TFGCL框架的预测性能更为稳健,较最优基线模型最高提升6.24%,TFGCL框架的稳健性尤其体现在数据缺失较为明显的数据集和长时交通流量预测任务中。

基于图自编码器和GRU网络的分层交通流预测模型

准确的交通流预测信息不仅可以为交通管理人员提供交通决策的坚实基础,还可以减少交通拥堵情况。在交通流预测任务中,获得有效的交通流的时空特性是保证预测效果的前提。现有的方法大多是用未来时刻的数据进行监督学习,提取的特征具有局限性。针对现有预测模型无法充分挖掘交通流的时空特性的问题,提出了基于改进的图自编码器和门控循环单元的分层交通预测模型。首先使用图注意力自编码器以无监督的方式深度挖掘交通流的空间特性,然后使用门控循环单元进行时间特征提取。分层结构采用分开训练的方式进行时空依赖关系的学习,旨在获取路网天然存在的空间拓扑特征,使其可以兼容不同时间步下的交通流预测任务。大量实验证明,所提出的GAE-GRU模型在不同数据集下的交通预测任务中取得了优异的表现,MAE,RMSE和MAPE指标均优于基线模型。

基于动态图神经常微分方程的地铁短时客流预测方法

随着城市轨道交通的快速发展,客流量的准确预测对于改善运营服务至关重要。为了解决当前地铁客流预测存在的时空特性挖掘不充分等问题,进一步提高预测的精度与效率,研究了基于动态图神经常微分方程模型(multivariate time series with dynamic graph neural ordinary differential equations,MTGODE)的地铁短时客流预测方法。该方法通彭颢1贺玉过学习地铁站点间的动态关联强度构建动态拓扑图结构,基于学习得到的动态图进行连续图传播以传递时空信息、挖掘客流的依赖关系,并采用残差卷积提取多时间尺度下的周期性模式,实现了对站点间时空动态的连续表征,克服了传统图卷积网络模型难以刻画动态空间依赖的局限性。此外,为了充分挖掘不同站点间客流分布的时空规律,综合利用北京地铁自动售检票系统(auto fare collection,AFC)刷卡数据、天气数据、空气质量数据以及车站周边用地属性数据构建多源融合的客流预测模型。通过选取地铁北京站和积水潭站-东直门站的历史数据开展进站客流和OD客流预测实验,结果表明:与多个基准模型相比,该模型在平均绝对误差、均方根误差和平均百分比误差这3个指标中均取得了更优的预测效果,相较最优基准模型扩散卷积循环神经网络(diffusion convolutional recurrent neural network,DCRNN)分别降低了9.93%,12.30%,9.23%,对地铁客流时空分布的拟合程度更好,模型具有更好的预测精度、稳定性和拟合能力。

自适应多视图融合图神经网络地铁客流预测模型

针对传统方法对地铁车站的多视角空间交互建模不足的问题,本文提出自适应多视图融合图神经网络模型(Adaptive Multi-view Fusion Graph Neural Network Model, AMFGNN)进行地铁车站短时客流预测。在空间维度,模型包括了物理拓扑图、线路可达性图、空间距离图等多个局部视图,并使用图注意力网络(Graph Attention Networks, GAT)学习每个视图内车站间的动态空间交互;以单视图车站为中心节点,结合其他视图中该车站作为邻居节点构建融合视图,并使用GAT学习多视图间动态交互;在时间维度,使用门控循环单元神经网络学习车站客流的时变特征。以重庆市地铁网络为例,全网出站客流的预测实验结果表明:相较于基线中的物理虚拟结合图网络模型(PVCGN),AMFGNN的平均绝对误差和均方根误差分别降低3.06%和2.49%。多视图内节点间注意力分数可视化结果表明,基于GAT的多视图建模思路能够自适应地融合不同视图中提取到的车站空间信息。此外,AMFGNN模型性能影响因素分析结果表明,以物理拓扑、线路可达性等结构稳定的视图作为中心节点构建融合视图能够获得更准确、稳定的预测模型。

基于双重存证的跨域流转多副本发现机制

针对泛在共享环境下信息频繁跨节点、跨系统交换时有意或无意留存而导致的隐私信息泄露问题,提出了一种基于双重存证的跨域流转多副本发现机制,可实现对流转信息的传播路径、传播方式进行溯源,并构造信息的多副本传播图。根据存证时机和方式的不同,双重存证包括流转主动存证和操作被动存证,信息在被分享前,由信息分享者主动记录传播路径和传播方式,生成流转主动存证记录;信息在被操作前,由系统自动记录传播路径,生成操作被动存证记录;相比单一存证,双重存证能够提高构造的信息多副本传播图的完整性和真实性,能够发现存证行为异常的节点并进行处置;基于社会惩戒理论,证明了存证行为异常发现与处置的有效性。实验开发了针对OFD的双重存证多副本发现原型系统,验证了所提机制对信息传播图构造完整性的提升。

多头图注意机制的深度学习地铁客流预测方法

挖掘地铁站点间的客流时空关联性是实现高精度地铁实时客流预测的关键。由于地铁网络中站点间客流的空间关联强弱性较为复杂且难以量化,从而导致客流预测过分依赖时间关联性。为克服该缺陷,提出一种多头图注意机制的深度学习地铁客流预测方法,通过构建面向地铁网络的多头图注意机制对多个关联站点间的客流空间关联性进行学习,得出差异化权重值,量化目标站点与关联站点群之间的客流空间关联强弱性;同时,将多头图注意机制融入长短时记忆学习模型,以量化后的空间关联性数据为输入,结合地铁客流量的时间关联性实现客流预测。实验结果表明,所提出的方法可行、有效,能够提升地铁客流预测精度且预测结果优于经典预测方法。

基于图论的两相渗流排序求解并行算法研究

对于两相渗流顺序求解格式中的饱和度计算,采用排序的方法可以将离散的全局非线性系统分解为一系列局部非线性系统来加速非线性求解过程.运用图论中的广度优先算法(BreadthFirstAlgorithm,BFS)和深度优先搜索(DepthFirst Search,DFS)来实现饱和度的并行计算。在两相流动方向一致时,基于BFS,使得同时参与计算的网格划分为一组,组与组之间满足拓扑排序的要求.在考虑重力和毛管力引起的两相流动方向不一致情况下,根据势能大小建立关于网格的有向有环图(DirectedCyclicGraph,DCG),利用强连通分量搜索的Tarjan算法找出耦合的网格,将耦合的网格视为一个网格并重新建图,从而得到有向无环图(Directed AcyclicGraph,DAG).最终实现饱和度排序求解的并行化.