Incremental clustering algorithm via crossentropy

A new incremental clustering method is presented, which partitions dynamic data sets by mapping data points in high dimension space into low dimension space based on （fuzzy） cross-entropy（CE）. This algorithm is divided into two parts： initial clustering process and incremental clustering process. The former calculates fuzzy cross-entropy or cross-entropy of one point relafive to others and a hierachical method based on cross-entropy is used for clustering static data sets. Moreover, it has the lower time complexity. The latter assigns new points to the suitable cluster by calculating membership of data point to existed centers based on the cross-entropy measure. Experimental compafisons show the proposed methood has lower time complexity than common methods in the large-scale data situations cr dynamic work environments.

Optical Fibre Communication Feature Analysis and Small Sample Fault Diagnosis Based on VMD-FE and Fuzzy Clustering

To solve the problems of a few optical fibre line fault samples and the inefficiency of manual communication optical fibre fault diagnosis,this paper proposes a communication optical fibre fault diagnosis model based on variational modal decomposition(VMD),fuzzy entropy(FE)and fuzzy clustering(FC).Firstly,based on the OTDR curve data collected in the field,VMD is used to extract the different modal components(IMF)of the original signal and calculate the fuzzy entropy(FE)values of different components to characterize the subtle differences between them.The fuzzy entropy of each curve is used as the feature vector,which in turn constructs the communication optical fibre feature vector matrix,and the fuzzy clustering algorithm is used to achieve fault diagnosis of faulty optical fibre.The VMD-FE combination can extract subtle differences in features,and the fuzzy clustering algorithm does not require sample training.The experimental results show that the model in this paper has high accuracy and is relevant to the maintenance of communication optical fibre when compared with existing feature extraction models and traditional machine learning models.

Research on Wind Power Prediction Modeling Based on Adaptive Feature Entropy Fuzzy Clustering

Wind farm power prediction is proposed based on adaptive feature weight entropy fuzzy clustering algorithm.According to the fuzzy clustering method,a large number of historical data of a wind farm in Inner Mongolia are analyzed and classified.Model of adaptive entropy weight for clustering is built.Wind power prediction model based on adaptive entropy fuzzy clustering feature weights is built.Simulation results show that the proposed method could distinguish the abnormal data and forecast more accurately and compute fastly.

Clustering Analysis of the Basic Structure of Relevant Community Service Organizations in Cities in China

With the gradually development of economy in China, people＇s living stan- dards have been improved, which makes people have higher and higher require- ments on the quality of life, and thus community service has become and essential part in people＇s life. In order to understand the basic building blocks of community service organizations in different cities in China, classification comparison was made to the data of 31 cities in China from China Statistical Year Book （2014） by using SPSS clustering method and the fuzzy clustering method, so as to find out the dif- ferences and the causes of the differences, with the aim to promote the manage- ment of relevant government and personnel.

Water quality assessment for Ulansuhai Lake using fuzzy clustering and pattern recognition

Water quality assessment of lakes is important to determine functional zones of water use.Considering the fuzziness during the partitioning process for lake water quality in an arid area,a multiplex model of fuzzy clustering with pattern recognition was developed by integrating transitive closure method,ISODATA algorithm in fuzzy clustering and fuzzy pattern recognition.The model was applied to partition the Ulansuhai Lake,a typical shallow lake in arid climate zone in the west part of Inner Mongolia,China and grade the condition of water quality divisions.The results showed that the partition well matched the real conditions of the lake,and the method has been proved accurate in the application.

AN ANALYSIS OF THE APPLICABILITY OF FUZZY CLUSTERING IN ESTABLISHING AN INDEX FOR THE EVALUATION OF METEOROLOGICAL SERVICE SATISFACTION

An evaluation index is a prerequisite for the scientific evaluation of a public meteorological service.This paper aims to explore a technical method for determining and screening evaluation indicators.Based on public satisfaction survey data obtained in Wafangdian,China in 2010,this study investigates the suitability of fuzzy clustering analysis method in establishing an evaluation index.Through quantitative analysis of multilayer fuzzy clustering of various evaluation indicators,correlation analysis indicates that if the results of clustering were identical for two evaluation indicators in the same sub-evaluation layer,then one indicator could be removed,or the two indicators merged.For evaluation indicators in different sub-evaluation layers,although clustering reveals attribute correlations,these indicators may not be substituted for one another.Analysis of the applicability of the fuzzy clustering method shows that it plays a certain role in the establishment and correction of an evaluation index.

A New Algorithm for Black-start Zone Partitioning Based on Fuzzy Clustering Analysis

On the process of power system black start after an accident, it can help to optimize the resources allocation and accelerate the recovery process that decomposing the power system into several independent partitions for parallel recovery. On the basis of adequate consideration of fuzziness of black-start zone partitioning, a new algorithm based on fuzzy clustering analysis is presented. Characteristic indexes are extracted fully and accurately. The raw data matrix is made up of the electrical distance between every nodes and blackstart resources. Closure transfer method is utilized to get the dynamic clustering. The availability and feasibility of the proposed algorithm are verified on the New-England 39 bus system at last.

TW-Co-MFC:Two-Level Weighted Collaborative Fuzzy Clustering Based on Maximum Entropy for Multi-View Data

In recent years,multi-view clustering research has attracted considerable attention because of the rapidly growing demand for unsupervised analysis of multi-view data in practical applications.Despite the significant advances in multi-view clustering,two challenges still need to be addressed,i.e.,how to make full use of the consistent and complementary information in multiple views and how to discriminate the contributions of different views and features in the same view to efficiently reveal the latent cluster structure of multi-view data for clustering.In this study,we propose a novel Two-level Weighted Collaborative Multi-view Fuzzy Clustering(TW-Co-MFC)approach to address the aforementioned issues.In TW-Co-MFC,a two-level weighting strategy is devised to measure the importance of views and features,and a collaborative working mechanism is introduced to balance the within-view clustering quality and the cross-view clustering consistency.Then an iterative optimization objective function based on the maximum entropy principle is designed for multi-view clustering.Experiments on real-world datasets show the effectiveness of the proposed approach.

EFFICIENT SUBSPACE CLUSTERING FOR HIGHER DIMENSIONAL DATA USING FUZZY ENTROPY

In this paper we propose a novel method for identifying relevant subspaces using fuzzy entropy and perform clustering. This measure discriminates the real distribution better by using membership functions for measuring class match degrees. Hence the fuzzy entropy reflects more information in the actual distribution of patterns in the subspaces. We use a heuristic procedure based on the silhouette criterion to find the number of clusters. The presented theories and algorithms are evaluated through experiments on a collection of benchmark data sets. Empirical results have shown its favorable performance in comparison with several other clustering algorithms.

Aggregation Similarity Measure Based on Hesitant Fuzzy Closeness Degree and Its Application to Clustering Analysis

In order to distinguish with effect different hesitant fuzzy elements(HFEs), we introduce the asymmetrical relative entropy between HFEs as a distance measure for higher discernment. Next,the formula of attribute weights is derived via an optimal model according to TOPSIS from the relative closeness degree constructed by the discerning relative entropy. Then, we propose the concept of cocorrelation degree from the viewpoint of probability theory and develop another new formula of hesitant fuzzy correlation coeffcient, and prove their similar properties to the traditional correlation coeffcient.To make full use of the existing similarity measures including the ones presented by us, we consider aggregation of similarity measures for hesitant fuzzy sets and derive the synthetical similarity formula.Finally, the derived formula is used for netting clustering analysis under hesitant fuzzy information and the effectiveness and superiority are veri?ed through a comparison analysis of clustering results obtained by other clustering algorithms.

ENTROPY TOLERANT FUZZY C-MEANS IN MEDICAL IMAGES

Segmenting the Dynamic Contrast-Enhanced Breast Magnetic Resonance Images(DCE-BMRI)is an extremely important task to diagnose the disease because it has the highest specificity when acquired with high temporal and spatial resolution and is also corrupted by heavy noise,outliers,and other imaging artifacts.In this paper,we intend to develop efficient robust segmentation algorithms based on fuzzy clustering approach for segmenting the DCE-BMRs.Our proposed segmentation algorithms have been amalgamated with effective kernel-induced distance measure on standard fuzzy c-means algorithm along with the spatial neighborhood information,entropy term,and tolerance vector into a fuzzy clustering structure for segmenting the DCE-BMRI.The significant feature of our proposed algorithms is its capability tofind the optimal membership grades and obtain effective cluster centers automatically by minimizing the proposed robust objective functions.Also,this article demonstrates the superiority of the proposed algorithms for segmenting DCE-BMRI in comparison with other recent kernel-based fuzzy c-means techniques.Finally the clustering accuracies of the proposed algorithms are validated by using silhouette method in comparison with existed fuzzy clustering algorithms.

基于VMD模糊熵与GG聚类的直流配电网故障检测方法

针对直流配电网存在的故障信号难以提取、不易对各类故障进行诊断等问题,提出一种基于变分模态分解(VMD)模糊熵与Gath-Geva(GG)聚类的故障检测方法。首先,提取出暂态电流,采用VMD算法将故障暂态电流分解成若干个固有模态分量(IMF)。然后,分别计算分解得到的若干个IMF的模糊熵,将其作为特征向量。最后,采用GG聚类算法对故障特征的特征向量进行聚类识别。GG聚类的主要算法为将聚类样本划分为c类,设出隶属度矩阵,通过设定迭代来计算聚类中心与最大似然估计距离,更新隶属度矩阵,当隶属度矩阵满足条件矩阵时终止迭代,从而实现对单极故障、极间故障以及区外交流侧接地故障的聚类识别。仿真结果表明,所提保护方案可靠性强、准确率高,在不同故障类型、故障位置和过渡电阻等工况下均能可靠检测直流线路故障并准确识别故障类型,且具备一定的抗干扰能力。

Track-Pattern-Based Characteristics of Extratropical Transitioning Tropical Cyclones in the Western North Pacific

Based on the Regional Specialized Meteorological Center(RSMC)Tokyo-Typhoon Center best-track data and the NCEP-NCAR reanalysis dataset,extratropical transitioning(ET)tropical cyclones(ETCs)over the western North Pacific(WNP)during 1951–2021 are classified into six clusters using the fuzzy c-means clustering method(FCM)according to their track patterns.The characteristics of the six hard-clustered ETCs with the highest membership coefficient are shown.Most tropical cyclones(TCs)that were assigned to clusters C2,C5,and C6 made landfall over eastern Asian countries,which severely threatened these regions.Among landfalling TCs,93.2%completed their ET after landfall,whereas 39.8%of ETCs completed their transition within one day.The frequency of ETCs over the WNP has decreased in the past four decades,wherein cluster C5 demonstrated a significant decrease on both interannual and interdecadal timescales with the expansion and intensification of the western Pacific subtropical high(WPSH).This large-scale circulation pattern is favorable for C2 and causes it to become the dominant track pattern,owning to it containing the largest number of intensifying ETCs among the six clusters,a number that has increased insignificantly over the past four decades.The surface roughness variation and three-dimensional background circulation led to C5 containing the maximum number of landfalling TCs and a minimum number of intensifying ETCs.Our results will facilitate a better understanding of the spatiotemporal distributions of ET events and associated environment background fields,which will benefit the effective monitoring of these events over the WNP.

考虑对称交互熵的对偶犹豫模糊企业环境行为决策模型分析

顺应绿色新发展理念,建设高质量生态文明,提升企业环境治理能力现代化水平,企业环境行为指标已逐渐被纳入投资者考量范围。本文针对第三方国际环境审计平台对企业环境行为评价的问题,研究面向对偶犹豫模糊评价信息的聚类方法。首先,定义对偶犹豫模糊相对熵与对称交互熵,并基于信息论角度提出一个新的对偶犹豫模糊相似度公式;然后,构造相似系数矩阵,并基于编网聚类方法对对偶犹豫模糊集进行聚类。最后,以不同指标权重下评价企业环境行为优劣程度为例进行计算,并与使用传统相关系数聚类方法所得结果进行对比,说明在考虑决策者提供的信息量较大的情况下,本文方法概念清晰、计算简单、分辨率高,是一种更加灵活、全面的多层次评价方法,为完善企业环境评价系统的构建提供了一种新思路。

随机正则3-可满足性问题的解簇结构分析

正则3-可满足性(3-SAT)问题是一个NP难问题,研究正则3-SAT问题解簇结构变化,旨在深入理解该问题的判定难度和可满足性解的分布情况。然而,现有分析模型只研究了接近簇集相变点的几个离散值,在不同约束密度下,缺乏统一的分析模型来描述解簇的结构演变。为了解决这一问题,提出解簇结构相变分析模型(PMSS)。该模型主要思想是采用WalkSAT算法和信息传播算法求得正则3-SAT问题可满足的初始解,再利用随机游走构造该初始解的解簇,并对解簇进行分析。用模块度和社区度量解簇社区结构,用结构熵度量解簇结构复杂性。实验结果表明,PMSS能够准确分析解簇结构演变过程,并且正则3-SAT问题实例的可满足相变点位于13~14,与使用Zchaff求解器得到的相变点一致,进一步验证了PMSS的有效性。

基于下肢sEMG的疲劳模糊增量熵表征方法研究

连续运动中,基于表面肌电信号(sEMG)外骨骼机器人与人进行协同运动控制,肌肉产生疲劳将影响人机协同控制的柔顺性及鲁棒性。本文创新性地提出模糊增量熵(EIFEn)用以表征肌肉疲劳程度,并对肌肉疲劳阶段的较为客观划分;采集人体连续抬腿运动中下肢12块肌肉的表面肌电信号,提出基于变异性敏感系数SVR肌肉疲劳敏感度判断方式,实现有效肌肉选取,提出基于均模积的自适应阈值动作切分法,将完整信号切分并提取单个动作信号序列,通过分析计算,对疲劳趋势进行表征。实验结果表明,本文模型相比时域频域算法具有较为明显的肌肉疲劳表征梯度特征,与fApEn及FFDispEn相比具有较好的疲劳表征能力,用于疲劳等级聚类的戴维森堡丁指数(DBI)为0.39,可提高外骨骼人机协同控制,为实现疲劳分阶段补偿助力提供参考。

可能性分布距离度量:一种鲁棒的域适应学习方法

领域适应(DA)学习旨在解决训练数据集与测试数据集分布不一致问题而广受关注,现有方法大多采用最小化领域间最大均值差(MMD)或其变体来解决域分布不一致问题。然而,领域中存在的噪声数据将会导致领域均值发生明显漂移,会在一定程度上影响基于MMD及其变体的学习方法的适应性能。故此,提出了可能性分布距离度量下的一种鲁棒的域适应学习方法:首先,将传统MMD准则变换为新颖的可能性聚类模型来削弱噪声数据所带来的影响,以此构建一种鲁棒的可能性分布距离度量(P-DDM)准则,并通过引入模糊熵正则项来进一步提升领域分布配准的鲁棒有效性。其次,基于P-DDM准则,提出一种鲁棒的域适应视觉分类机(C-PDDM),其引入图拉普拉斯矩阵来保留源域与目标域内部数据间的几何结构一致性,以提升标签传播性能,同时通过最大化利用源域判别信息进行最小化领域判别误差,以进一步提升学习模型的泛化性能。理论分析证实,在一定条件下,所提P-DDM是传统分布距离度量方法MMD准则的一个上界,因而通过最小化P-DDM能有效优化MMD目标。最后,与几个代表性的领域适应学习方法进行比较,在6个视觉基准数据集(Office31、Office-Caltech、Office-Home、PIE、MNIST-UPS和COIL20)上的实验结果显示,该方法在泛化性能上平均提升了5%左右,在鲁棒性能上平均提升了10%左右。

模糊熵与灰云模型在既有玻璃幕墙安全评价中的应用

针对既有玻璃幕墙的安全评价问题,考虑指标信息的不确定性与等级归属的复杂性,提出一种基于模糊熵与灰云聚类模型的既有玻璃幕墙安全评价方法。首先,以实际工程为背景,通过识别既有玻璃幕墙安全风险因素,构建一套定性与定量指标相结合的安全评价指标体系;其次,基于可变模糊云与正态云模型进行评价指标的赋值;再次,运用极差最大化组合赋权法,计算评价指标的权重;最后,引入灰云聚类模型与模糊熵,实现既有玻璃幕墙安全状态的二维评估。实例结果显示:该方法既能对玻璃幕墙的安全状态进行诊断,还能运用模糊熵,从复杂性的层面揭示既有幕墙的健康状况,进而为既有玻璃幕墙的安全评估提供可行思路。

城市轨道交通线网线路客流热度评价研究

为评估新时期城市轨道交通线网客流快速回升的情况,从基础客流、基础线网、出行特征以及运力与运能4个方面选取12项评价指标,对北京市轨道交通线网的客流热度进行深入研究。采用熵权优劣解距离法(TOPSIS)和聚类分析方法,对北京市轨道交通路网中的15条线路客流进行客流热度评价和分类。研究结果显示,10号、1-八通、6号和5号线作为大环线或贯通线,均属于高热度客流线路。而除11号线外,其余线路均属于中热度客流线路,呈现出明显的区域性客流特点。基于这些发现,提出以下建议:北京市轨道交通发展应更关注中低热度地区;运营企业可根据线路客流热度的不同,合理分配人力和物力资源;一线部门可根据客流热度的预测提前采取措施应对客流冲击。

基于ITD模糊熵和GG聚类的滚动轴承故障诊断

提出了一种本征时间尺度分解模糊熵和GG模糊聚类的滚动轴承故障诊断方法。首先,将滚动轴承的振动信号进行ITD分解,得到若干个固有旋转分量和一个趋势项。然后,将PR分量分别与原始信号进行相关性分析,筛选出前3个含主要特征信息的PR分量,并将筛选的PR分量的模糊熵作为特征向量。最后,将特征向量输入到GG分类器中进行聚类识别。通过模糊熵、样本熵和近似熵对比,实验结果表明模糊熵能更好的表征故障信号的特征信息;通过GG聚类、GK聚类和FCM聚类对比,实验结果表明GG聚类效果明显优于FCM、GK的聚类效果。因此,实验证明了基于ITD模糊熵和GG聚类的滚动轴承故障诊断方法的有效性和优越性。