基于CWT-CNN的滚动轴承故障诊断

针对滚动轴承传统故障诊断方法训练收敛速度慢、识别准确率不高、抗噪性能差等问题,提出CWT-CNN的轴承故障诊断模型。通过对滚动轴承振动数据经连续小波变换生成的时频图进行三次垂直方向随机裁剪的方法扩充数据集,之后将其导入到搭建的加入了批量归一化和随机失活的卷积神经网络中进行模型训练,再由训练好的模型实现轴承故障分类。为了测试模型性能,使用凯斯西储大学轴承数据集进行检测,经过实验结果表明：基于提出的方法构建的数据集相比于常规方法,在搭建的卷积神经网络训练中收敛速度更快,训练出的模型性能也更加稳定,最终最高测试准确率为99.75%,常规方法构建的数据集准确率为99.67%,证明了构建数据集方法的可行性;在原始数据中加入信噪比为6dB高斯白噪声后,通过常规方法构建的数据集测试的最高准确率仍达到了98.67%,展现了基于CWT-CNN的轴承故障诊断模型较强的抗噪能力,证明了所提方法的有效性和可行性。     

一种新的沥青路面灌封裂缝自动提取方法

为了提高基于图像处理的沥青路面病害识别效率和精度,引入图像增强处理中的多尺度视网膜(multi-scale Retinex, MSR)算法以减弱光照不均匀、道路场景多变等因素对路面病害图像质量的影响。针对SegNet网络难以精确分割沥青路面微小病害的问题,采用比视觉几何群网络(visual geometry group network, VGG)效果更好的残差网络(residual network, ResNet)作为主干网络,同时加入空洞卷积(dilation convolution)层,提高网络对细小病害的识别性能;针对改进网络在识别病害时误检率较高的问题,运用阈值法剔除分割结果中的假阳性。为了验证改进算法的有效性,将其与具有代表性的语义分割方法(如SegNet、BiSeNet)在相同数据集上进行对比,三者的平均交并比(mean intersection over union, MIoU)分别为0.776 3、0.674 3、0.697 1,三者的F₁分数(F₁-score,F₁)分别为0.899 9、0.874 3...     

基于自注意力机制的新生儿胆道闭锁识别

新生儿胆道闭锁是新生儿常见的致命疾病之一,并且该病在亚洲的发病率高于世界其他地区.新生儿胆道闭锁需要及时发现及时治疗,然而由于缺少专业的儿科医生和辅助诊疗手段,新生儿父母往往不能及时发现而错过了最佳治疗时间.因此,本文开发了一个具有实际应用价值的预诊算法,通过新生儿粪便图片预测新生儿是否患有新生儿胆道闭锁,并提醒新生儿父母及时就诊.为了让算法在应用场景下识别率更高,本文的算法基于一个真实场景下拍摄的新生儿粪图片数据集开发.首先我们设计了一个自注意力网络模型BANet(Biliary Atresia Network),将图片的浅层特征和深层特征相结合,可以得到更好的分类效果.由于拍摄自应用场景下的图片存在过暗和过曝等问题.通过分析数据集的亮度分布,我们设计了一个自动亮度调节算法解决.此外,图片中的阴影也会对识别结果造成干扰,因此我们在训练阶段增加了一种阴影数据增强方式来缓解这一问题.为验证本文提出算法的有效性,本文设计了一个和医生的对比试验.结果证明BANet在四分类的识别率、二分类的识别率、特异性和敏感性等客观评价指标上占有明显优势.本文提出的BANet能够有效利用图片中的颜色、异常点...     

Image Retrieval Based on Objects＇ Spatial Orientation Relationships

A new concept of characteristic scanning radial (CSR) is proposed for the segmented image on the basis of two shape-speciiic points oi its shape-objects. Subsequently, two characteristic attribute sequences (CAS) of relative distance and relative direction are derived to represent the spatial orientation relationships among objects of the image. A novel image retrieval algorithm is presented using these two CASs. The proposed retrieval approach not only satisfies the transformational invariance, but also attains the quantitative comparison of matching. Experiments identify the effectiveness and efficiency of the algorithm adequately.     

Incremental Web Usage Mining Based on Active Ant Colony Clustering

To alleviate the scalability problem caused by the increasing Web using and changing users＇ interests, this paper presents a novel Web Usage Mining algorithm-Incremental Web Usage Mining algorithm based on Active Ant Colony Clustering. Firstly, an active movement strategy about direction selection and speed, different with the positive strategy employed by other Ant Colony Clustering algorithms, is proposed to construct an Active Ant Colony Clustering algorithm, which avoid the idle and ＂flying over the plane＂ moving phenomenon, effectively improve the quality and speed of clustering on large dataset. Then a mechanism of decomposing clusters based on above methods is introduced to form new clusters when users＇ interests change. Empirical studies on a real Web dataset show the active ant colony clustering algorithm has better performance than the previous algorithms, and the incremental approach based on the proposed mechanism can efficiently implement incremental Web usage mining.     

Combined Viterbi detector for a balanced code in page memories

Based on the two path metrics being equal at a merged node in the trellis employed to describe a Viterbi detector for the detection of data encoded with a rate 6 : 8 balanced binary code in page-oriented optical memories, the combined Viterbi detector scheme is proposed to improve raw biterror rate performance by mitigating the occurrence of a twobit reversing error event in an estimated codeword for the balanced code. The effectiveness of the detection scheme is verified for different data quantizations using Monte Carlo simulations.     

Orientl: A strategy of Web information integration

We propose the OrientI approach for effectively building flexible applications on information integration. The system presents a fully visual development environment tobuild applications. With Orient I system, a user only needs to concentrate on the composition of components for building the InterPlan. and the detailed underlying operations and data streams are invisible to the user. A prototype system has been implemented and has partially proved the convenience hrought by the Orient] approach.     

A distributed dynamic clustering algorithm for wireless sensor networks

This paper proposes a distributed dynamic k-medoid clustering algorithm for wireless sensor networks (WSNs), DDKCAWSN. Different from node-clustering algorithms and protocols for WSNs, the algorithm focuses on clustering data in the network. By sending the sink clustered data instead of practical ones, the algorithm can greatly reduce the size and the time of data communication, and further save the energy of the nodes in the network and prolong the system lifetime. Moreover, the algorithm improves the accuracy of the clustered data dynamically by updating the clusters periodically such as each day. Simulation results demonstrate the effectiveness of our approach for different metrics. Biography: WANG Leichun (1974–), male, Ph.D. candidate, research direction: wireless communication.     

New Abnormal Cervical Cell Detection Method of Multi-Spectral Pap Smears

Considering the problem of traditional cervical cancer detection method that brings high false negative rate （FNR） and high false positive rate （FPR）, a new abnormal cervical cells detection method of multi-spectral Pap smear is proposed in this thesis, on the basis of multi-spectral microscopic imaging technology and computer automotive recognition technology. At first, image in a specific wave band is segmented according to the relationship between intensity and spectrum of each pixel. Then, multi-spectral features of each pixel are extracted making use of improved cosine correlation analysis （CCA） algorithm. Combined with the characteristic of each cell＇s area, final definition is made. Experiments have proved the new approach could identify abnormal cells efficiently as well as lower FNR and FPR.     

Fault Tolerant Approach for Data Encryption and Digital Signature Based on ECC System

An integrated fault tolerant approach for data encryption and digital signature based on elliptic curve cryptography is proposed. This approach allows the receiver to verify the sender‘s identity and can simultaneously deal with error detection and data correction. Up to three errors in our approach can be detected and corrected. This approach has atleast the same security as that based on RSA system, but smaller keys to achieve the same level of security. Our approach is more efficient than the known ones and more suited for limited environments like personal digital assistants (PDAs), mobile phones and smart cards without RSA coprocessors.     

基于深度卷积神经网络的三维模型识别

为了进一步提高三维模型的识别精度,提出了一种基于深度卷积神经网络的三维模型识别方法。将点云数据通过占用网格规范化计算转化为二值3D体素矩阵,通过附加正则化项的随机梯度下降算法提取体素矩阵的特征,再通过共享权重的旋转增强对训练集进行数据增广并以此对模型标签进行预测。实验结果表明,该算法在公开数据集ModelNet40及悉尼城市模型数据集上的识别精度均达到85%左右。与基于同类机器学习的三维模型识别算法相比,在相同训练数据集上该方法网络训练时间短,在相同测试数据集上模型识别准确率高,检索速度快。提出的体素占用网格模型的深度卷积神经网络,可以实现三维点云模型数据集及规范化体素模型数据集的识别和分类工作。     

A Partially Non-Cryptographic Security Routing Protocol in Mobile Ad Hoc Networks

In this paper, we propose a partially non-cryptographic security routing protocol （PNCSR） that protects both routing and data forwarding operations through the same reactive approach. PNCSR only apply public-key cryptographic system in managing token, but it doesn＇t utilize any cryptographic primitives on the routing messages. In PNCSR, each node is fair. Local neighboring nodes collaboratively monitor each other and sustain each other. It also uses a novel credit strategy which additively increases the token lifetime each time a node renews its token. We also analyze the storage, computation, and communication overhead of PNCSR, and provide a simple yet meaningful overhead comparison. Finally, the simulation results show the effectiveness of PNCSR in various situations.     

Multi-Dimensional Customer Data Analysis in Online Auctions

In this paper, we designed a customer-centered data warehouse system with five subjects： listing, bidding, transaction, accounts, and customer contact based on the business process of online auction companies. For each subject, we analyzed its fact indexes and dimensions. Then take transaction subject as example, analyzed the data warehouse model in detail, and got the multi-dimensional analysis structure of transaction subject. At last, using data mining to do customer segmentation, we divided customers into four types： impulse customer, prudent customer, potential customer, and ordinary customer. By the result of multi-dimensional customer data analysis, online auction companies can do more target marketing and increase customer loyalty.     

基于T检验与支持向量机的蛋白质质谱数据分析

对蛋白质质谱数据进行模式识别成为癌症诊断的一种新方法,但质谱数据存在高维小样本问题,因而数据分析面临着巨大挑战.在对原始数据进行基线校正与标准化并用分箱法进行降维预处理的基础上,提出用T检验方法选取特征,对蛋白质质谱数据进行分析研究.实验对卵巢质谱数据集进行分类,用10-fold交叉验证法选择训练和测试样本,以支持向量...     

基于表情及姿态融合的情绪识别

情绪识别指在使计算机拥有能够感知和分析人类情绪和意图的能力,从而在娱乐、教育、医疗和公共安全等领域发挥作用.与直观的面部表情相比,身体姿态在情绪识别方面的作用总是被低估.针对公共空间个体人脸分辨率较低、表情识别精度不高的问题,提出了融合面部表情和身体姿态的情绪识别方法.首先,对视频数据进行预处理获得表情通道和姿态通道的输入序列;然后,使用深度学习的方法分别提取表情和姿态的情绪特征;最后,在决策层进行融合和分类.构建了基于视频的公共空间个体情绪数据集(SCU-FABE),在此基础上,结合姿态情绪识别数据增强,实现了公共空间个体情绪的有效识别.实验结果表明,表情和姿态情绪识别取得了94.698%和88.024%的平均识别率;融合情绪识别平均识别率为95.766%,有效融合了面部表情和身体姿态表达的情绪信息,在真实场景视频数据中具有良好的泛化能力和适用性.     

基于深度卷积神经网络与中心损失的人脸识别

传统人脸识别方法手工设计特征过程复杂、识别率较低,对于开集人脸识别通用深度学习分类模型特征判别能力较弱。针对这两方面的不足,提出了一种以分类损失与中心损失相结合作为模型训练监督信号的深度卷积神经网络。首先,利用构建的应用场景数据集优调从公共数据集获得初始化参数的深度人脸识别模型,解决训练数据过小和数据分布差异问题,同时提高模型训练速度;然后,以传统损失函数和新的中心损失作为迁移学习过程中的监督信号,使得类内聚合、类间分散,提高模型输出人脸特征的判别能力;最后,对人脸特征进行主成分分析,进一步去除冗余特征,降低特征复杂度,提高人脸识别准确率。实验结果表明,与传统人脸识别算法相比该算法可以自动进行特征提取,并且相对于通用深度学习分类模型该算法通过度量学习使特征表示更具判别力。在自建测试集和LFW、YouTube Faces标准测试集上都取得了较高的识别率。     

A Scalable Intrusion Detection System for IPv6

0 IntroductionIPv6 ,as a new, widely available version of Internet proto-col , contains numerous features that make it attractivefroma securitystandpoint[1].IPv6is not a panaceafor securi-ty,though, because the basic mechanisms for transportingpackets across the network stay mostly unchanged,and theupper-layer protocols that transport the actual application dataare mostly unaffected.IPv6 is usually deployed without cryp-tographic protections of any kind in real environment[2].So,IPv6 securi…     

Implementation of an Autocorrelation Pitch Detector in Application to Query by Humming

A pitch detector for application in query by humming (QBH) is implemented in this paper. This algorithm is made up of two parts: note segmentation and pitch detection. In the first part, voiced/silence decision is made on each segment of the input signal by a pattern recognition approach, and further, the preparatory note borders are obtained; then, via analysis of the instantaneous energy contour, the adjacent notes that adhere to each other are separated. In the second part, pitch is estimated for all frames contained in a note‘s duration by an autocorrelation method and the mean of these pitch values is taken as the average pitch of this note. Moreover, in order to remove the effect of formant structure, a nonlinear preprocessing is adopted in the pitch detection part and the autocorrelation function is properly weighted before peak picking. Finally, hummings of several experimenters with different voice characters are recorded to test this pitch detector, whose efficiency and reliability are proved by the result.     

2013～2017年医院尿路感染住院患者抗菌药物应用分析

目的 通过调查医院2013~2017年尿路感染住院患者抗菌药物使用情况,为评价临床药师参与临床用药中发挥的作用提供依据。方法 检索宜春市人民医院2013~2017年456例尿路感染住院患者病例资料,分析比较2013~2015与2016~2017年抗菌药物的使用与患者治疗情况。包括用药频度（DDDs）、平均使用品种、联合用药率、换药率、选药不合理率、剂量不合理率、疗程不合理率、平均住院天数及治愈率。结果 2013~2015年氟喹诺酮类药物DDDS排序均为第一;2016~2017年头孢菌素类药物DDDS排序均为第一;单类品种中,左氧氟沙星DDDs排序在2013~2016年期间为第一;2017年时氨曲南DDDS排序位居首位而左氧氟沙星位居第五;2013~2017年患者平均用药品种数与疗程不合理率无显著变化,联合用药率、换药率于2016~2017年较以往有明显降低P<0.01,P<0.05,但平均住院天数较往年有明显升高P<0.05。剂量不合理率、选药不合理率较往年降低而治愈率较往年有所提高。结论 在临床药师的参与下,尿路感染患者抗菌药物用药逐渐走向更加规范合理,尿路感染患者的抗菌药物应用得到有效控制。