UAV data link anti-interference via SLHS-SVM-AdaBoost algorithm:Classification prediction and route planning

The ability to predict the anti-interference communications performance of unmanned aerial vehicle(UAV)data links is critical for intelligent route planning of UAVs in real combat scenarios.Previous research in this area has encountered several limitations:Classifiers exhibit low training efficiency,their precision is notably reduced when dealing with imbalanced samples,and they cannot be applied to the condition where the UAV’s flight altitude and the antenna bearing vary.This paper proposes the sequential Latin hypercube sampling(SLHS)-support vector machine(SVM)-AdaBoost algorithm,which enhances the training efficiency of the base classifier and circumvents local optima during the search process through SLHS optimization.Additionally,it mitigates the bottleneck of sample imbalance by adjusting the sample weight distribution using the AdaBoost algorithm.Through comparison,the modeling efficiency,prediction accuracy on the test set,and macro-averaged values of precision,recall,and F1-score for SLHS-SVM-AdaBoost are improved by 22.7%,5.7%,36.0%,25.0%,and 34.2%,respectively,compared with Grid-SVM.Additionally,these values are improved by 22.2%,2.1%,11.3%,2.8%,and 7.4%,respectively,compared with particle swarm optimization(PSO)-SVM-AdaBoost.Combining Latin hypercube sampling with the SLHS-SVM-AdaBoost algorithm,the classification prediction model of anti-interference performance of UAV data links,which took factors like three-dimensional position of UAV and antenna bearing into consideration,is established and used to assess the safety of the classical flying path and optimize the flying route.It was found that the risk of loss of communications could not be completely avoided by adjusting the flying altitude based on the classical path,whereas intelligent path planning based on the classification prediction model of anti-interference performance can realize complete avoidance of being interfered meanwhile reducing the route length by at least 2.3%,thus benefiting both safety and operation efficiency.

Research on Heterogeneous Information Network Link Prediction Based on Representation Learning

A heterogeneous information network,which is composed of various types of nodes and edges,has a complex structure and rich information content,and is widely used in social networks,academic networks,e-commerce,and other fields.Link prediction,as a key task to reveal the unobserved relationships in the network,is of great significance in heterogeneous information networks.This paper reviews the application of presentation-based learning methods in link prediction of heterogeneous information networks.This paper introduces the basic concepts of heterogeneous information networks,and the theoretical basis of representation learning,and discusses the specific application of the deep learning model in node embedding learning and link prediction in detail.The effectiveness and superiority of these methods on multiple real data sets are demonstrated by experimental verification.

Link prediction based on a semi-local similarity index

Missing link prediction provides significant instruction for both analysis of network structure and mining of unknown links in incomplete networks. Recently, many algorithms have been proposed based on various node-similarity measures. Among these measures, the common neighbour index, the resource allocation index, and the local path index, stemming from different source, have been proved to have relatively high accuracy and low computational effort. In this paper, we propose a similarity index by combining the resource allocation index and the local path index. Simulation results on six unweighted networks show that the accuracy of the proposed index is higher than that of the local path one. Based on the same idea of the present index, we develop its corresponding weighted version and test it on several weighted networks. It is found that, except for the USAir network, the weighted variant also performs better than both the weighted resource allocation index and the weighted local path index. Due to the improved accuracy and the still low computational complexity, the indices may be useful for link prediction.

Entropy-based link prediction in weighted networks

Information entropy has been proved to be an effective tool to quantify the structural importance of complex networks.In a previous work [Xu et al. Physica A, 456 294（2016）], we measure the contribution of a path in link prediction with information entropy. In this paper, we further quantify the contribution of a path with both path entropy and path weight,and propose a weighted prediction index based on the contributions of paths, namely weighted path entropy（WPE）, to improve the prediction accuracy in weighted networks. Empirical experiments on six weighted real-world networks show that WPE achieves higher prediction accuracy than three other typical weighted indices.

Link prediction in complex networks via modularity-based belief propagation

Link prediction aims at detecting missing, spurious or evolving links in a network, based on the topological information and/or nodes＇ attributes of the network. Under the assumption that the likelihood of the existence of a link between two nodes can be captured by nodes＇ similarity, several methods have been proposed to compute similarity directly or indirectly, with information on node degree. However, correctly predicting links is also crucial in revealing the link formation mechanisms and thus in providing more accurate modeling for networks. We here propose a novel method to predict links by incorporating stochastic-block-model link generating mechanisms with node degree. The proposed method first recov- ers the underlying block structure of a network by modularity-based belief propagation, and based on the recovered block structural information it models the link likelihood between two nodes to match the degree sequence of the network. Experiments on a set of real-world networks and synthetic networks generated by stochastic block model show that our proposed method is effective in detecting missing, spurious or evolving links of networks that can be well modeled by a stochastic block model. This approach efficiently complements the toolbox for complex network analysis, offering a novel tool to model links in stochastic block model networks that are fundamental in the modeling of real world complex networks.

Combat network link prediction based on embedding learning

Link prediction of combat networks is of significant military value for precisely identifying the vital infrastructure of the enemy target and optimizing the operational plan of our side.Due to the profound uncertainty in the battleground circumstances, the acquired topological information of the opponent combat network always presents sparse characteristics. To solve this problem, a novel approach named network embedding based combat network link prediction(NECLP) is put forward to predict missing links of sparse combat networks. First,node embedding techniques are presented to preserve as much information of the combat network as possible using a low-dimensional space. Then, we put forward a solution algorithm to predict links between combat networks based on node embedding similarity. Last, massive experiments are carried out on a real-world combat network case to verify the validity and practicality of the proposed NECLP. This paper compares six baseline methods, and experimental results show that the NECLP has outstanding performance and substantially outperforms the baseline methods.

Knowledge Graph Representation Learning Based on Automatic Network Search for Link Prediction

Link prediction,also known as Knowledge Graph Completion(KGC),is the common task in Knowledge Graphs(KGs)to predict missing connections between entities.Most existing methods focus on designing shallow,scalable models,which have less expressive than deep,multi-layer models.Furthermore,most operations like addition,matrix multiplications or factorization are handcrafted based on a few known relation patterns in several wellknown datasets,such as FB15k,WN18,etc.However,due to the diversity and complex nature of real-world data distribution,it is inherently difficult to preset all latent patterns.To address this issue,we proposeKGE-ANS,a novel knowledge graph embedding framework for general link prediction tasks using automatic network search.KGEANS can learn a deep,multi-layer effective architecture to adapt to different datasets through neural architecture search.In addition,the general search spacewe designed is tailored forKGtasks.We performextensive experiments on benchmark datasets and the dataset constructed in this paper.The results show that our KGE-ANS outperforms several state-of-the-art methods,especially on these datasets with complex relation patterns.

Cold-Start Link Prediction via Weighted Symmetric Nonnegative Matrix Factorization with Graph Regularization

Link prediction has attracted wide attention among interdisciplinaryresearchers as an important issue in complex network. It aims to predict the missing links in current networks and new links that will appear in future networks.Despite the presence of missing links in the target network of link prediction studies, the network it processes remains macroscopically as a large connectedgraph. However, the complexity of the real world makes the complex networksabstracted from real systems often contain many isolated nodes. This phenomenon leads to existing link prediction methods not to efficiently implement the prediction of missing edges on isolated nodes. Therefore, the cold-start linkprediction is favored as one of the most valuable subproblems of traditional linkprediction. However, due to the loss of many links in the observation network, thetopological information available for completing the link prediction task is extremely scarce. This presents a severe challenge for the study of cold-start link prediction. Therefore, how to mine and fuse more available non-topologicalinformation from observed network becomes the key point to solve the problemof cold-start link prediction. In this paper, we propose a framework for solving thecold-start link prediction problem, a joint-weighted symmetric nonnegative matrixfactorization model fusing graph regularization information, based on low-rankapproximation algorithms in the field of machine learning. First, the nonlinear features in high-dimensional space of node attributes are captured by the designedgraph regularization term. Second, using a weighted matrix, we associate the attribute similarity and first order structure information of nodes and constrain eachother. Finally, a unified framework for implementing cold-start link prediction isconstructed by using a symmetric nonnegative matrix factorization model to integrate the multiple information extracted together. Extensive experimental validationon five real networks with attributes shows that the proposed model has very goodpredictive performance when predicting missing edges of isolated nodes.

Semantic-aware graph convolution network on multi-hop paths for link prediction

Knowledge graph(KG) link prediction aims to address the problem of missing multiple valid triples in KGs. Existing approaches either struggle to efficiently model the message passing process of multi-hop paths or lack transparency of model prediction principles. In this paper,a new graph convolutional network path semantic-aware graph convolution network(PSGCN) is proposed to achieve modeling the semantic information of multi-hop paths. PSGCN first uses a random walk strategy to obtain all-hop paths in KGs,then captures the semantics of the paths by Word2Sec and long shortterm memory(LSTM) models,and finally converts them into a potential representation for the graph convolution network(GCN) messaging process. PSGCN combines path-based inference methods and graph neural networks to achieve better interpretability and scalability. In addition,to ensure the robustness of the model,the value of the path thresholdKis experimented on the FB15K-237 and WN18RR datasets,and the final results prove the effectiveness of the model.

Doubly-Fed Pumped Storage Units Participation in Frequency Regulation Control Strategy for New Energy Power Systems Based on Model Predictive Control

Large-scale new energy grid connection leads to the weakening of the system frequency regulation capability,and the system frequency stability is facing unprecedented challenges.In order to solve rapid frequency fluctuation caused by new energy units,this paper proposes a new energy power system frequency regulation strategy with multiple units including the doubly-fed pumped storage unit(DFPSU).Firstly,based on the model predictive control(MPC)theory,the state space equations are established by considering the operating characteristics of the units and the dynamic behavior of the system;secondly,the proportional-differential control link is introduced to minimize the frequency deviation to further optimize the frequency modulation(FM)output of the DFPSU and inhibit the rapid fluctuation of the frequency;lastly,it is verified on theMatlab/Simulink simulation platform,and the results show that the model predictive control with proportional-differential control link can further release the FM potential of the DFPSU,increase the depth of its FM,effectively reduce the frequency deviation of the system and its rate of change,realize the optimization of the active output of the DFPSU and that of other units,and improve the frequency response capability of the system.

ADAPTIVE PREDICTIVE CONTROL OF NEAR-SPACE VEHICLE USING FUNCTIONAL LINK NETWORK

A novel nonlinear adaptive control method is presented for a near-space hypersonic vehicle （NHV） in the presence of strong uncertainties and disturbances. The control law consists of the optimal generalized predictive controller （OGPC） and the functional link network （FLN） direct adaptive law. OGPC is a continuous-time nonlinear predictive control law. The FLN adaptive law is used to offset the unknown uncertainties and disturbances in a flight through the online learning. The learning process does not need any offline training phase. The stability analyses of the NHV close-loop system are provided and it is proved that the system error and the weight learning error are uniformly ultimately hounded. Simulation results show the satisfactory performance of the con- troller for the attitude tracking.

Adaptive nonlinear model predictive control design of a flexible-link manipulator with uncertain parameters

This paper presents a novel adaptive nonlinear model predictive control design for trajectory tracking of flexible-link manipulators consisting of feedback linearization, linear model predictive control, and unscented Kalman filtering. Reducing the nonlinear system to a linear system by feedback linearization simplifies the optimization problem of the model predictive controller significantly, which, however, is no longer linear in the presence of parameter uncertainties and can potentially lead to an undesired dynamical behaviour. An unscented Kalman filter is used to approximate the dynamics of the prediction model by an online parameter estimation, which leads to an adaptation of the optimization problem in each time step and thus to a better prediction and an improved input action. Finally, a detailed fuzzy-arithmetic analysis is performed in order to quantify the effect of the uncertainties on the control structure and to derive robustness assessments. The control structure is applied to a serial manipulator with two flexible links containing uncertain model parameters and acting in three-dimensional space.

基于多特征提取和对比学习的知识图谱链接预测

针对传统知识图谱链接预测方法提取图谱节点特征角度单一,且在训练过程中较少考虑节点间复杂的交互作用,构建的负例三元组质量较低等问题,提出了一种链接预测方法,旨在充分利用知识图谱节点间的相互作用和图结构蕴含的交互信息,考虑从多特征角度识别出三元组中的缺失事实。首先,通过不同的节点特征提取方式从不同角度获得节点的嵌入表示,并聚合邻居节点特征以增强其实体语义信息;其次,用多个卷积操作提取实体和关系之间的全局关系和过渡特征,通过深度特征提取的方式处理实体和关系的信息交互;最后,通过引入对比学习,干预负例三元组的构建,同时增强负例三元组的特征,提高所构建三元组的质量,最终通过计算余弦相似度筛选出预测实体。实验结果表明,提出的方法在知识图谱链接预测任务中的多个评价指标相比对比模型均有提高,同时验证了所提方法在处理多关系的复杂知识图谱时的有效性。

用于在线社交网络的链路预测好友推荐算法JAFLink

现在的社交网络不只是人们现实生活社交圈的一个反映,同时也在一定程度上扩展着人们的交际范围,使得用户在社交网上找到更多适合自己的朋友.但是,由于社交网络发展迅速,用户量巨大,对于用户来说,自己从中找到自己的好友是比较困难的,这就需要社交网站向用户提供一个比较好的推荐算法,从而使得网站真正能够改变用户的生活.本文提出的JAFLink(Jaccad-Adamic Adar-Feature)链路加权方法,结合jaccad和Adamic Adar并考虑了社交网络用户属性,构成JAFLink加权方法,用来计算节点对建立链接的可能性.实验表明,JAFLink相比只考虑网络拓扑结构或者用户属性而言更加高效.

基于边扰动的链接预测解释方法

多数链接预测模型是解释性较差的黑盒模型,因此不少学者提出了针对链接预测的解释方法,但这些方法存在着解释的目标模型单一、缺乏泛化能力、解释结果准确率不足等缺陷。为弥补这些不足,提出一种基于边扰动的链接预测的解释方法。首先利用广度优先搜索得到从头实体到尾实体的所有路径,随后搜索路径所经过实体的邻居节点,形成待解释三元组的训练子图;然后采用边扰动的方式在训练子图上重新训练嵌入模型,计算每条边对预测结果的影响程度;最后通过双向的束搜索得到对预测结果影响程度最大的路径,作为待解释三元组的解释路径。实验表明,该方法在公共数据集上的性能超过了大多数的链接预测解释方法,ACC相较于最先进的方法提升了2.3%,AUPR提升了1.9%。同时在生物医学数据集上针对使用链接预测技术的药物重定位任务进行结果的解释实验,其解释体现了良好的可理解性、启发性。提出了一种不依赖于特定模型且有效的解释方法,该方法通过边扰动和路径搜索得到解释路径,使结果的解释更加直观和易于理解,同时能够为不同领域的知识图谱应用提供支持。

单样本条件下邻域选择聚合零次知识图谱链接预测方法

为了解决支持样本有限条件下零次知识图谱链接预测模型性能下降的问题,提出了一种单样本条件下邻域选择聚合零次知识图谱链接预测方法(NSALP)。该方法主要由特征提取器、生成器、判别器三个模块实现。借鉴图同构网络的思想对特征提取器模块进行改进,在聚合头尾邻域时为每个邻域节点分配一个可学习的参数,进而过滤无关特征,凸显有效特征;以头节点嵌入与关系文本描述的组合作为生成器学习过程的引导,使生成器生成的新组合特征更加接近真实的知识三元组结构特征。在NELL-ZS和Wiki-ZS两个零次知识图谱数据集上,所提模型的性能对比基线模型分别提升了2.5和0.7百分点。在NELL-ZS进行的消融实验中,所提extractor+和generator+模块的性能表现均优于未做改进的模型,佐证了改进方法的有效性。

基于特征学习的链路预测模型TNTlink

在合作作者网络中,链路预测可以预测当前网络中缺失的链接,以及新的或已解散的链接,根据网络中观测到的信息来推断两位作者在不久的将来是否会产生合作,对于挖掘和分析网络的演化、重塑网络模型具有重要意义。链路预测是计算机科学和物理学的重要研究方向,对此已有较深入的研究,其主要研究思路是基于马尔可夫链、机器学习和无监督的学习。然而,这些工作大多只使用单一的特征,即基于网络拓扑特征或者属性特征进行预测,很少将这些跨学科的特征组合考虑,结合多学科特征进行链路预测的研究非常少。文中设计开发了TNTlink模型,该模型结合网络拓扑特征、基本特征和附加特征,并结合物理学和计算机科学的领域知识,利用深度神经网络将这些特征集成到一个深度学习框架中,其在解决链路预测问题时取得了不错的效果。文中使用了5个数据集(ca-AstroPh,ca-CondMat,ca-GrQc,ca-HepPh和ca-HepTh),包含69032个节点和450617条边,从捕获的信息中利用二进制相似度和模糊余弦相似度计算和识别特征。如果节点在这些特征中表现出更多的相似性(如相似的节点、相同的关键字或彼此之间密切的关系),则两个节点间更有可能生成链接。除了考虑节点的特征外,还考虑了节点重要性对链路形成的影响,进而提出了一种新的链路预测指标MI,以区分强影响和弱影响,对节点的重要影响进行建模。将所提模型与主流分类器在5个数据集上进行比较,结果表明MI和TNTlink有效地提高了链路预测的AUC值。

多元关系知识表示学习方法研究综述

知识表示学习旨在将知识库中的实体和关系转化为机器能够理解和处理的形式,从而提升模型的分析与推理能力。针对传统二元关系知识表示学习的局限,如忽略高阶关系、缺乏扩展性和有限的表达力,多元关系知识表示学习方法应运而生。全面综述了多元关系知识表示学习方法。梳理和分析了知识表示学习相关综述工作;阐释了知识表示学习和链接预测的基本概念,并根据超图、角色、超关系这三种表示形式,定义了多元关系知识表示学习任务;从基于平移距离、张量分解、卷积神经网络、图神经网络和其他类型五类方法,展示了该领域的研究进展;介绍了常用的数据集与评价指标,并通过链接预测任务评估了不同模型的性能;探讨了目前方法存在的问题和挑战,并对未来的研究方向提出了展望。

融合关系和结构编码的规则抽取与推理研究

领域知识图谱拥有不完备性和语义复杂多样性的特点,从而导致其在规则抽取和选择问题上的不足,影响了其推理的能力.针对此问题,提出了一种融合关系和结构编码的规则抽取模型.通过提取目标子图中的关系和结构信息并进行特征编码,从而实现了一种多维度的嵌入表达方法.设计了融合关系和结构信息的自注意力机制,使模型能够更好地捕捉输入序列中的依赖关系和局部结构信息,从而提升了模型对于上下文的理解和表达能力,进而解决了在语义复杂情况下规则的抽取和选择的问题.通过在真实汽车部件故障工业数据集和公共数据集的实验,表明了在链接预测与规则质量评估任务中,所提出的模型都有一定的提升(规则长度为3时, mean reciprocal rank (MRR)平均提升了7.1百分点, Hits@10平均提升了8.6百分点;规则长度为2时, MRR平均提升了7.4百分点, Hits@10平均提升了3.9百分点),证实了关系和结构信息对于规则抽取与推理的有效性.

多模态特征增强的双层融合知识推理方法

现有的多模态知识推理方法大多采用拼接或注意力的方式,将预训练模型提取到的多模态特征直接进行融合,往往忽略了不同模态之间的异构性和交互的复杂性。为此,提出了一种多模态特征增强的双层融合知识推理方法。结构信息嵌入模块采用自适应图注意力机制筛选并聚合关键的邻居信息,用来增强实体和关系嵌入的语义表达;多模态嵌入信息模块使用不同的注意力机制关注不同模态数据的独有特征,以及多模态数据间的共性特征,利用共性特征的互补信息进行模态交互,以减少模态间异构性差异;多模态特征融合模块采用将低秩多模态特征融合和决策融合相结合的双层融合策略,实现了多模态数据在模态间和模态内的动态复杂交互,并综合考虑每种模态在推理中的贡献度,得到更全面的预测结果。为了验证方法的有效性,分别在FB15K-237、DB15K和YAGO15K数据集上进行了实验。结果表明:该方法相比多模态推理方法,在FB15K-237数据集上MRR和Hits@1分别平均提升3.6%和2.2%;相比单模态推理方法,MRR和Hits@1分别平均提升13.7%和14.6%。