An Improved Directed Acyclic Graph Support Vector Machine

In this paper, we propose an improved Directed Acyclic Graph Support Vector Machine （DAGSVM） for multi-class classification. Compared with the traditional DAGSVM, the improved version has advantages that the structure of the directed acyclic graph is not chosen random and fixed, and it can be adaptive to be optimal according to the incoming testing samples, thus it has a good generalization performance. From experiments on six datasets, we can see that the proposed improved version of DAGSVM is better than the traditional one with respect to the accuracy rate.

Multi-Class Support Vector Machine Classifier Based on Jeffries-Matusita Distance and Directed Acyclic Graph

Based on the framework of support vector machines （SVM） using one-against-one （OAO） strategy, a new multi-class kernel method based on directed aeyclie graph （DAG） and probabilistic distance is proposed to raise the multi-class classification accuracies. The topology structure of DAG is constructed by rearranging the nodes＇ sequence in the graph. DAG is equivalent to guided operating SVM on a list, and the classification performance depends on the nodes＇ sequence in the graph. Jeffries-Matusita distance （JMD） is introduced to estimate the separability of each class, and the implementation list is initialized with all classes organized according to certain sequence in the list. To testify the effectiveness of the proposed method, numerical analysis is conducted on UCI data and hyperspectral data. Meanwhile, comparative studies using standard OAO and DAG classification methods are also conducted and the results illustrate better performance and higher accuracy of the orooosed JMD-DAG method.

Thrust estimator design based on least squares support vector regression machine

In order to realize direct thrust control instead of traditional sensor-based control for aero-engines,it is indispensable to design a thrust estimator with high accuracy,so a scheme for thrust estimator design based on the least square support vector regression machine is proposed to solve this problem. Furthermore,numerical simulations confirm the effectiveness of our presented scheme. During the process of estimator design,a wrapper criterion that can not only reduce the computational complexity but also enhance the generalization performance is proposed to select variables as input variables for estimator.

Structural Reliability Analysis Based on Support Vector Machine and Dual Neural Network Direct Integration Method

Aiming at the reliability analysis of small sample data or implicit structural function,a novel structural reliability analysis model based on support vector machine(SVM)and neural network direct integration method(DNN)is proposed.Firstly,SVM with good small sample learning ability is used to train small sample data,fit structural performance functions and establish regular integration regions.Secondly,DNN is approximated the integral function to achieve multiple integration in the integration region.Finally,structural reliability was obtained by DNN.Numerical examples are investigated to demonstrate the effectiveness of the present method,which provides a feasible way for the structural reliability analysis.

BOOSTING SPARSE LEAST SQUARES SUPPORT VECTOR REGRESSION (BSLSSVR) AND ITS APPLICATION TO THRUST ESTIMATION

In order to realize direct thrust control instead of conventional sensors-based control for aero-engine, a thrust estimator with high accuracy is designed by using the boosting technique to improve the performance of least squares support vector regression （LSSVR）. There exist two distinct features compared with the conven- tional boosting technique： （1） Sampling without replacement is used to avoid numerical instability for modeling LSSVR. （2） To realize the sparseness of LSSVR and reduce the computational complexity, only a subset of the training samples is used to construct LSSVR. Thus, this boosting method for LSSVR is called the boosting sparse LSSVR （BSLSSVR）. Finally, simulation results show that BSLSSVR-based thrust estimator can satisfy the requirement of direct thrust control, i.e. , maximum absolute value of relative error of thrust estimation is not more than 5‰.

A Fast Algorithm for Support Vector Clustering

Support Vector Clustering (SVC) is a kernel-based unsupervised learning clustering method. The main drawback of SVC is its high computational complexity in getting the adjacency matrix describing the connectivity for each pairs of points. Based on the proximity graph model [3], the Euclidean distance in Hilbert space is calculated using a Gaussian kernel, which is the right criterion to generate a minimum spanning tree using Kruskal's algorithm. Then the connectivity estimation is lowered by only checking the linkages between the edges that construct the main stem of the MST (Minimum Spanning Tree), in which the non-compatibility degree is originally defined to support the edge selection during linkage estimations. This new approach is experimentally analyzed. The results show that the revised algorithm has a better performance than the proximity graph model with faster speed, optimized clustering quality and strong ability to noise suppression, which makes SVC scalable to large data sets.

基于BiLSTM-LSSVM的螺杆转子铣削加工廓形预测

针对螺杆转子盘铣刀加工过程中的轮廓预测问题,提出了基于双向长短时神经网络-最小二乘支持向量机(BiLSTM-LSSVM)的螺杆廓形预测方法。首先,对加工过程中的振动信号进行采集并进行降噪预处理,降噪后的信号进行降采样处理随后输入BiLSTM中进行时序预测;其次,对时序预测后的信号进行特征提取,将提取后的特征向量输入LSSVM进行廓形预测;最后,以五头螺杆为例通过正交实验对BiLSTM-LSSVM模型进行试验验证,并对预测廓形进行误差补偿实验。实验结果表明,提出的基于BiLSTM-LSSVM的螺杆廓形预测模型可对螺杆转子盘铣刀加工螺杆廓形进行准确预测,进而为螺杆转子加工廓形补偿提供支持。

基于时频图学习的北斗卫星导航系统干扰类型识别

针对当前干扰识别算法依赖特征工程提取特征存在提取繁琐、识别准确率受干信比取值影响较大的问题,本文提出在不同干信比情况下基于时频图学习的北斗卫星导航系统(BDS,Beidou navigation satellite system)干扰类型识别方法。以航空机载北斗卫星导航系统B1I信号为对象,对原始B1I信号、包含干扰的B1I信号进行短时傅里叶变换,将变换后获取的时频图作为支持向量机和卷积神经网络模型的输入向量,完成干扰类型的检测与识别。仿真结果表明,两种机器学习识别算法的平均识别率均达到了99%以上,识别结果比传统决策树识别算法提升了30%以上,解决了现有干扰识别算法需要严重依赖人工设计的特征工程手动提取干扰信号特征以及识别率较低的问题。该研究结果可为后续的干扰抑制工作提供先验信息,提高航空领域中北斗卫星导航系统的安全性。

一种海上风电场柔直送出线单端保护方案

远海、大容量的海上风电场需要通过高压柔性直流系统并入大陆交流主网,柔直线路运行于海底,故障率高,需要配置快速、可靠的直流线路保护。目前,已实际应用于柔直工程的单端行波保护的暂态量幅值在高阻故障场景下会明显降低,其保护性能有进一步提升的空间。故障行波在传播过程中经过限流电抗,行波幅值变化受到抑制导致行波波形具有平滑特征,而区内故障下故障行波波形较曲折,突变明显。根据此特征,本文提出一种基于主成分分析(PCA)和支持向量机(SVM)的柔性直流线路单端保护方法,经仿真实验表明,该保护方法具有较好的耐过渡电阻及抗干扰能力。

基于多层支持向量机的蓄电池在线监测方法研究

常规的蓄电池在线监测节点一般采用独立或区域布设形式,监测的范围受限制,导致在线监测频率均值下降,因此探讨了基于多层支持向量机的蓄电池在线监测方法。根据当前测定,先进行蓄电池电动势及开路电压应用数值采集,采用自适应的方式,打破监测范围的限制,部署自适应在线监测节点;然后构建多层支持向量机的蓄电池在线监测模型,采用持续跟踪预警处理强化在线监测。针对选定的4个蓄电池,按照顺序分别植入0.8 mm、1.2 mm及2.1 mm的4个型号的电阻丝,形成不同的电阻率。经测定计算最终得出在线监测频率均值均可达到150 Hz以上,说明设计的蓄电池在线监测方法更加稳定、安全,在不同环境下的整体适应度更强,监测效率得到了显著提升。

Single-ended Fault Detection Scheme Using Support Vector Machine for Multi-terminal Direct Current Systems Based on Modular Multilevel Converter

This paper proposes a single-ended fault detection scheme for long transmission lines using support vector machine(SVM)for multi-terminal direct current systems based on modular multilevel converter(MMC-MTDC).The scheme overcomes existing detection difficulties in the protection of long transmission lines resulting from high grounding resistance and attenuation,and also avoids the sophisticated process of threshold value selection.The high-frequency components in the measured voltage extracted by a wavelet transform and the amplitude of the zero-mode set of the positive-sequence voltage are the inputs to a trained SVM.The output of the SVM determines the fault type.A model of a four-terminal DC power grid with overhead transmission lines is built in PSCAD/EMTDC.Simulation results of EMTDC confirm that the proposed scheme achieves 100%accuracy in detecting short-circuit faults with high resistance on long transmission lines.The proposed scheme eliminates mal-operation of DC circuit breakers when faced with power order changes or AC-side faults.Its robustness and time delay are also assessed and shown to have no perceptible effect on the speed and accuracy of the detection scheme,thus ensuring its reliability and stability.

Detection and recognition of LPI radar signals using visibility graphs

The detection and recognition of radar signals play a critical role in the maintenance of future electronic warfare(EW).So far,however,there are still problems with signal detection and recognition,especially in the low probability of intercept(LPI)radar.This paper explores the usefulness of such an algorithm in the scenario of LPI radar signal detection and recognition based on visibility graphs(VG).More network and feature information can be extracted in the VG two-dimensional space,this algorithm can solve the problem of signal recognition using the autocorrelation function.Wavelet denoising processing is introduced into the signal to be tested,and the denoised signal is converted to the VG domain.Then,the signal detection is performed by using the constant false alarm of the VG average degree.Next,weight the converted graph.Finally,perform feature extraction on the weighted image,and use the feature to complete the recognition.It is testified that the proposed algorithm offers significant improvements,such as robustness to noise,and the detection and recognition accuracy,over the recent researches.

Automatic radar antenna scan type recognition based on limited penetrable visibility graph

To address the problem of the weak anti-noise and macro-trend extraction abilities of the current methods for identifying radar antenna scan type,a recognition method for radar antenna scan types based on limited penetrable visibility graph(LPVG)is proposed.Firstly,seven types of radar antenna scans are analyzed,which include the circular scan,sector scan,helical scan,raster scan,conical scan,electromechanical hybrid scan and two-dimensional electronic scan.Then,the time series of the pulse amplitude in the radar reconnaissance receiver is converted into an LPVG network,and the feature parameters are extracted.Finally,the recognition result is obtained by using a support vector machine(SVM)classifier.The experimental results show that the recognition accuracy and noise resistance of this new method are improved,where the average recognition accuracy for radar antenna type is at least 90%when the signalto-noise ratio(SNR)is 5 dB and above.

Electromagnetic side-channel attack based on PSO directed acyclic graph SVM

Machine learning has a powerful potential for performing the template attack(TA) of cryptographic device. To improve the accuracy and time consuming of electromagnetic template attack(ETA), a multi-class directed acyclic graph support vector machine(DAGSVM) method is proposed to predict the Hamming weight of the key. The method needs to generate K(K ? 1)/2 binary support vector machine(SVM) classifiers and realizes the K-class prediction using a rooted binary directed acyclic graph(DAG) testing model. Further, particle swarm optimization(PSO) is used for optimal selection of DAGSVM model parameters to improve the performance of DAGSVM. By exploiting the electromagnetic emanations captured while a chip was implementing the RC4 algorithm in software, the computation complexity and performance of several multi-class machine learning methods, such as DAGSVM, one-versus-one(OVO)SVM, one-versus-all(OVA)SVM, Probabilistic neural networks(PNN), K-means clustering and fuzzy neural network(FNN) are investigated. In the same scenario, the highest classification accuracy of Hamming weight for the key reached 100%, 95.33%, 85%, 74%, 49.67% and 38% for DAGSVM, OVOSVM, OVASVM, PNN, K-means and FNN, respectively. The experiment results demonstrate the proposed model performs higher predictive accuracy and faster convergence speed.

不同材质绝缘子污秽等级高光谱检测方法研究

该文提出了适用于不同材质(陶瓷、玻璃和硅橡胶)绝缘子表面污秽等级的高光谱检测方法,采集不同材质和污秽等级样本的高光谱数据,经预处理后提取标签集及检测集谱线数据;采用玻璃材质样本标签集数据建立污秽等级检测模型;并运用分段直接标准化校正陶瓷、硅橡胶样本谱线数据,实现同一模型下不同材质样本的污秽等级检测。结果表明:不同材质样本同一污秽等级下,高光谱谱线吸收峰、反射峰位置及变化趋势有明显差异;同一材质不同污秽等级谱线差异主要为幅值。检测模型对玻璃、陶瓷和硅橡胶样本的污秽等级检测准确率分别为98.3%、95.0%和91.7%,并利用人工积污试验对模型进行了验证,污秽等级检测准确率为83.3%,证明了该模型可有效实现不同材质绝缘子表面污秽等级的高光谱检测。

小波检测和特征图谱决策的非侵入电动自行车充电实时监测系统

电动自行车违规入户充电行为具有时间随机性以及空间隐蔽性,存在较大安全隐患且难以有效管理。利用非侵入式监测系统具有实时自主执行和便捷易推广的特性,文中提出了基于小波检测和特征图谱决策的非侵入式电动自行车充电实时监测系统。考虑电动自行车负荷的物理结构和充电特性,从暂态和稳态两方面分析电动自行车负荷的典型共性特征;预先构建具有强可分性和通用性的电动自行车专有特征图谱实现电动自行车稳态共性特征的一致性结构化表征;实际监测过程中,为了降低系统的算力需求和数据传输压力,基于小波变换精确定位具有高频分量的电动自行车专有暂态现象完成类电动自行车充电事件检测。最后,提取事件波形并通过图谱训练高效分类器进行负荷认定并实时上传。通过对实际用户进行监测,验证了监测系统的有效性,可以有效解决电动自行车进楼入户充电的问题。

基于KG-DBN-SVM的工控网络安全态势感知算法

工业控制网络是重要的基础设施,保障其安全稳定运行非常重要。对工控网络进行安全态势感知研究,可以帮助安全人员从更加全面的层面发现潜在威胁,保障工控网络安全。工控网络数据来源很多、结构各异,存在多源异构的特点,从这一点出发对数据进行分析,可以更好地感知工控网络安全态势。本文使用知识图谱对多源异构数据进行结构化,然后利用深度置信网络对不同工控实体数据进行特征提取与降维,最后利用支持向量机进行分类判断确定,并进行数据调优,得到最佳的工控网络安全态势感知模型。在公共的工控安全数据集上进行对比实验,实验结果表明,本文算法在准确率、召回率与F1指标上分别达到了0.938、0.891和0.914的结果,优于对比较的一系列工控网络安全算法。

脑功能网络拓扑属性改变支持向量机模型预测海马硬化型颞叶癫痫患者术后转归

目的基于海马硬化(HS)型颞叶癫痫(TLE)患者脑功能网络拓扑属性改变建立支持向量机(SVM)模型,观察其预测HS-TLE患者术后转归的价值。方法回顾性收集34例因单侧HS-TLE接受前颞叶切除术(ATL)患者(HS-TLE组)及50名健康对照者(HC组);根据术后癫痫Engel分级将HS-TLE患者分为无癫痫发作(SF)亚组(EngelⅠa级,n=20)及癫痫发作(NSF)亚组(EngelⅠb~Ⅳ级,n=14)。基于头部静息态功能MRI(rs-fMRI)图论分析组间及HS-TLE组内亚组间大脑网络节点拓扑属性的差异;基于亚组间存在显著差异的图论指标构建SVM模型预测HS-TLE患者术后转归,并评估其预测效能。结果相比HC组,HS-TLE组患侧海马介数中心性降低,额下回岛盖部、额下回眶部等度中心性升高而双侧海马等度中心性降低,额下回三角部节点效率升高而海马、海马旁回及杏仁核节点效率降低。HS-TLE组内,相比NSF亚组,SF亚组患侧杏仁核及对侧直回、角回、颞中回介数中心性降低,而健侧枕上回、梭状回介数中心性升高;健侧枕上回和枕中回度中心性及节点效率均升高。SVM模型预测HS-TLE患者ATL术后转归的准确率为76.47%(26/34)。结论脑功能网络拓扑属性变化SVM模型可用于预测HS-TLE患者ATL术后转归。

基于MSST及HOG特征提取的雷达辐射源信号识别

针对传统雷达信号识别算法在低信噪比下识别准确率低的问题,提出了基于多重同步压缩(MSST)时频变换及方向梯度直方图(HOG)特征提取的雷达辐射源信号识别算法。所提算法在雷达时域信号短时傅里叶变换(STFT)基础上进行多重同步压缩处理获得信号时频分布图,通过HOG算子对信号时频分布图进行HOG特征提取,将提取的HOG特征通过主成分分析法(PCA)进行降维,将降维后的特征参数送入支持向量机(SVM)对雷达信号进行分类与识别。实验结果表明:所提算法具有较低的复杂度,当信噪比为-8 dB时,仿真实验与半实物仿真实验针对9种典型雷达信号的识别准确率达到90%以上。

基于难治性癫痫患者脑网络特征的立体脑电图引导射频热凝毁损术预后预测

射频热凝毁损术(RFTC)治疗难治性癫痫的疗效的个体差异较大。本课题旨在研究术前脑网络图论指标,建立预测RFTC预后的模型。基于45例难治性癫痫患者术前的立体脑电图数据,建立时变多变量自回归模型,通过计算频谱加权的时变部分指向性相干,构建时变效应连接网络,计算图论指标。根据RFTC后至少3个月的Engel分级,将患者分为RFTC有效组(EngelⅠ和Ⅱ级)与RFTC无效组(EngelⅢ级),进行组间图论指标的统计学分析,并基于图论指标,应用支持向量机(SVM)建模进行预后预测。结果表明,RFTC有效组的标准化的平均聚类系数(P=0.000)、小世界性(P=0.022)显著高于RFTC无效组,标准化的特征路径长度显著低于RFTC无效组(P=0.032)(RFTC有效组的标准化的平均聚类系数、小世界性和标准化的特征路径长度分别为0.9953±0.0002、0.8530±0.0062和1.1688±0.0085;RFTC无效组的标准化的平均聚类系数、小世界性和标准化的特征路径长度分别为0.9940±0.0002、0.8335±0.0056和1.1944±0.0080);应用以上3个指标,通过SVM进行RFTC疗效的预测,准确率达到81.97%。应用难治性癫痫患者术前的脑效应连接网络图论指标标准化的平均聚类系数、标准化的特征路径长度和小世界性建立的预后预测模型可以很好地预测RFTC疗效。