Modeling of wire electro-discharge machining of TiC/Fe in situ metal matrix composite using normalized RBFN with enhanced k-means clustering technique

In the present work, wire electro-discharge machinability of 5 vol% TiC/Fe in situ metal matrix composite (MMC) has been studied. Four input process parameters such as pulse on-time, pulse off-time, wire feed-rate, and average gap voltage have been considered, while cutting speed and kerf width have been considered as the measure of performance of the process. The presence of nonconductive TiC particles and formation of Fe₂O₃ during machining make the process very much unstable and stochastic. Thus, modeling the process either by an analytical or numerical method becomes extremely difficult. In the present study, modeling of wire electro-discharge machining process by normalized radial basis function network (NRBFN) with enhanced k-means clustering technique has been done. In order to measure the effectiveness of this approach, the process has also been modeled by NRBFN with traditional k-means technique, and a comparison has been made between the two models. It is seen that both the models can predict the cutting speed and kerf width successfully, but NRBFN with enhanced k-means clustering technique yields better results than NRBFN with traditional k-means technique. Both the models have been used to carry out the parametric study and, finally, have been compared with the experimental results.     

迭代可调节网络聚类及其在故障诊断中的应用

针对常用聚类算法对复杂分布数据难以有效聚类的问题,把网络分析技术与基于代价函数最优的聚类技术相结合,提出一种新颖的迭代可调节网络聚类算法。该算法采用网络的思想建立样本空间模型,把数据聚类问题转化为基于节点生长连接的网络分析问题;并设计了可调节的节点间相似关系测度和相应的聚类准则来构建节点间邻域搜索及节点生长操作;通过改变调节系数来实现网络节点间连接关系的整体调节。新算法能够在无需预先设定簇数目的情况下,自动获得簇的数目和样本数据的分布位置。采用4组不同样本分布的人工数据集聚类和往复压缩机气阀泄漏故障诊断试验,对比测试了新算法与K均值算法(KM)的性能,结果表明迭代可调节网络聚类算法可实现对复杂分布的流形数据聚类,在准确率及自动处理程度性能指标上明显优于常用的KM算法。     

基于谱聚类的振动多模态信号幅谱分割研究与应用

在强噪声下频域的模态峰往往受到强烈的干扰,导致模态参数的提取精度下降,甚至产生模态主频误判。针对这种情形,采用谱聚类算法对振动频谱进行宏观聚类,提出了一种新的幅谱分割方法。按照波峰概念把振动信号幅谱分割成波峰的集合,把每个波峰看成一个待聚类的样本,构建波峰相似度函数、拉普拉斯矩阵和聚类算法,引入谱聚类算法进行波峰自动聚类,聚类的结果就是宏观上的单模态大峰。仿真试验表明,这种幅谱波峰分割的谱聚类算法能够减小噪声和虚假模态的影响,与已有的k-means聚类算法相比,具有更强的噪声抵抗能力和更好的聚类能力。通过对斜拉索振动进行模态测试,证实该算法能够得到符合肉眼观察的幅谱分割效果,且具有较好的稳定性和准确性。     

Structural damage detection by fuzzy clustering

The development of strategies for structural health monitoring (SHM) has become increasingly important because of the necessity of preventing undesirable damage. This paper describes an approach to this problem using vibration data. It involves a three-stage process: reduction of the time-series data using principle component analysis (PCA), the development of a data-based model using an auto-regressive moving average (ARMA) model using data from an undamaged structure, and the classification of whether or not the structure is damaged using a fuzzy clustering approach. The approach is applied to data from a benchmark structure from Los Alamos National Laboratory, USA. Two fuzzy clustering algorithms are compared: fuzzy c-means (FCM) and Gustafson–Kessel (GK) algorithms. It is shown that while both fuzzy clustering algorithms are effective, the GK algorithm marginally outperforms the FCM algorithm.     

Psoriasis image identification using k-means clustering with morphological processing

We present a preliminary design and experimental results of psoriasis objects tracking method for color-skin images that utilizes k-means clustering with morphological processing technique. The method is capable of solving unable exactly contoured psoriasis objects problem in color-skin image by adding the morphological reconstruction operation. The key idea of the proposed image processing procedure is the k-means clustering method helps the rough segmentation, then the dilation and erosion method are adapted to refine previous results. In this paper we investigate the possibility of employing this approach for psoriasis image application. The application of the proposed method for tracking psoriasis is demonstrated to help pathologists distinguish exactly its size and region. In this paper, we propose a psoriasis image segmentation procedure to improve the accuracy. The experimental results demonstrate that the misclassification error is very small between the proposed result and hand drawing.     

Recursive fuzzy c-means clustering for recursive fuzzy identification of time-varying processes

In this paper we propose a new approach to on-line Takagi-Sugeno fuzzy model identification. It combines a recursive fuzzy c-means algorithm and recursive least squares. First the method is derived and than it is tested and compared on a benchmark problem of the Mackey-Glass time series with other established on-line identification methods. We showed that the developed algorithm gives a comparable degree of accuracy to other algorithms. The proposed algorithm can be used in a number of fields, including adaptive nonlinear control, model predictive control, fault detection, diagnostics and robotics. An example of identification based on a real data of the waste-water treatment process is also presented.     

多视图K-多均值聚类算法

多均值聚类算法假设每个类拥有多个子类,通过求解优化问题的方式来求解每个样本子类的划分和最终类簇的划分.该算法弥补了K-均值算法在非球数据集上的劣势,取得了较好的聚类效果,但是该算法无法被运用到多视图数据集上.本文提出了一种多视图K-多均值聚类算法,保留了K-多均值设置多个子类的设计,引入了视图权重参数,将目标聚类数作为限制条件,通过求解最优问题获得最终的类簇.将本文提出的算法与流行的多视图聚类算法进行对比实验,证明了本文算法的优越性.     

A differential evolution fuzzy clustering approach to machine cell formation

Searching for similar parts and associated machines plays a crucial role in machine cell design. The aim of this paper is to propose a new fuzzy clustering method to solve machine cell formation problems in a fuzzy environment. Fuzzy c-means (FCM) has been applied to machine cell formation problems. However, performance of existing FCM algorithms depends highly on the initial states (cluster centers or membership degrees). Additionally, the number of clusters (machine cells) has to be provided beforehand. In practice, it is difficult for the machine cell designer to select a proper set of initial states or to determine the optimal number of machine cells before the overall machine cell configuration is formed and the operational result is observed. In this paper, a new fuzzy clustering approach combining differential evolution (DE) algorithms with FCM formulas is proposed to overcome these deficiencies. The proposed DE-based fuzzy clustering method can automatically determine the correct number of cells and generate an optimal machine cell configuration at the same time. Experimental results demonstrate that the proposed algorithm performs well in searching solutions to the fuzzy machine cell formation problem with automatic cluster number determination.     

基于异构值差度量的SOM混合属性数据聚类算法

针对传统聚类算法处理混合属性数据聚类质量不高且聚类结果可视化差的问题,提出了基于异构值差度量的自组织映射混合属性数据聚类算法。该算法以自组织映射神经网络为框架,采用基于样本概率的异构值差度量混合属性数据的相异性。利用分类特征项在Voronoi集合中出现频率作为分类属性数据参考向量更新规则的基础,通过混合更新规则实现数值属性和分类属性数据规则的更新。利用UCI公共数据库中的分类属性和混合属性数据集来测试所提出的聚类算法,并与SOM算法和kprototypes、SBAC、KL-FCM-GM算法进行比较。最后将所提出的聚类算法应用于轮式移动机器人的运动状态分析,获得了较好的聚类效果。     

A novel KFCM based fault diagnosis method for unknown faults in satellite reaction wheels

Reaction wheels are one of the most critical components of the satellite attitude control system, therefore correct diagnosis of their faults is quintessential for efficient operation of these spacecraft. The known faults in any of the subsystems are often diagnosed by supervised learning algorithms, however, this method fails to work correctly when a new or unknown fault occurs. In such cases an unsupervised learning algorithm becomes essential for obtaining the correct diagnosis. Kernel Fuzzy C-Means (KFCM) is one of the unsupervised algorithms, although it has its own limitations; however in this paper a novel method has been proposed for conditioning of KFCM method (C-KFCM) so that it can be effectively used for fault diagnosis of both known and unknown faults as in satellite reaction wheels. The C-KFCM approach involves determination of exact class centers from the data of known faults, in this way discrete number of fault classes are determined at the start. Similarity parameters are derived and determined for each of the fault data point. Thereafter depending on the similarity threshold each data point is issued with a class label. The high similarity points fall into one of the 'known-fault' classes while the low similarity points are labeled as 'unknown-faults'. Simulation results show that as compared to the supervised algorithm such as neural network, the C-KFCM method can effectively cluster historical fault data (as in reaction wheels) and diagnose the faults to an accuracy of more than 91%.     

利用高斯混合模型的多光谱图像模糊聚类分割

针对传统分割算法难以实现高分辨率多光谱图像分割的问题,本文提出一种利用高斯混合模型的多光谱图像模糊聚类分割算法。该算法采用高斯混合模型定义像素对类属的非相似性测度,由于该算法具有高精度拟合数据统计分布能力,故可以有效剔除噪声对分割结果的影响。同时,引入隐马尔科夫随机场(Hidden Markov Random Field,HMRF)定义邻域作用的先验概率,并将其作为各高斯分量权值以及KL(Kullback-Leibler)信息中控制聚类尺度的参数,从而增强了算法对复杂场景遥感图像的鲁棒性,进一步提高了算法的分割精度。对模拟图像和高分辨多光谱图像分割结果进行了定性定量分析。实验结果表明:模拟图像的总精度达96.8%以上。这验证了本文算法在分割高分辨率多光谱图像时具有保留细节信息的能力,而且也证实了算法的有效性和可行性。该算法能够实现高分辨率多光谱图像的精确分割。     

A typical process route discovery method based on clustering analysis

A typical process route is a sample of planning the process route. It is a kind of the process planning knowledge. In order to discover the typical process route in the process planning database from the Computer Aided Process Planning (CAPP), Knowledge Discovery in Database (KDD) is applied. Process data selection, process data purge and process data transformation are employed to get optimized process data. The clustering analysis is adopted as the algorithm mining the typical process route. A mathematics model describing the process route was built by the data matrix. There are three similarities in process route clustering: the similarity between operations was measured by the Manhattan distance based on operation code; the similarity between process routes was calculated by the Euclidean distance and expressed as a dissimilarity matrix; the similarity between process route clusters was evaluated by the average distance based on the dissimilarity matrix. Then, the process route clusters were eventually merged by the agglomerative hierarchical clustering method. And the process routes clustering result was determined by the clustering granularity of process route. This method has been applied successfully to discovering the typical process route of a kind of axle sleeves. This project is supported by the National High-Tech. R&D Program for CIMS, China (Grant No. 2003AA411041).     

混合聚类新算法及其在故障诊断中的应用

针对模糊C-均值(FCM)聚类算法假设各维特征和每个样本对聚类贡献相同,同时需要预先设定聚类数的不足,利用3层前馈神经网络、点密度函数算法和聚类有效性指标对其进行改进,提出一种新的混合聚类算法。该算法考虑到不同特征和不同样本对聚类结果有不同程度的影响,并根据聚类有效性指标的变化自适应确定聚类数来实现聚类。利用基于梯度下降的3层前馈神经网络通过无监督训练来自适应学习特征权值,使用基于点密度函数的算法获取样本权值,给不同特征和不同样本赋予权重,突出敏感特征和典型样本的主导作用,抑制其他特征和样本对聚类的干扰,以提高聚类性能。研究结果表明,对于国际标准测试数据和某机车轴承的早期故障诊断,该混合聚类算法不但能自动确定聚类数,而且聚类的准确性明显比FCM高。     

基于核熵成分分析的流式数据自动分群方法

针对多参数流式细胞数据传统人工分群过程复杂、自动化程度不高等问题,提出了一种基于核熵成分分析(KECA)的自动分群方法。选取对瑞利(Renyi)熵具有最大贡献的特征向量作为投影方向,对数据进行特征提取;设计了一种基于余弦相似度和K-means算法的分类器,并采用一种基于向量夹角的最佳聚类数确定方法,最终获得细胞的分类标签。对实验获得的淋巴细胞免疫表型分析数据进行处理,结果表明,该方法能够实现细胞的快速、自动分群,整体分群准确率能够达到97%以上,操作简单便捷,提高了细胞分析的效率。     

改进的FCM聚类算法

经典FCM聚类算法存在的两个方面的问题：一是算法对初始聚类中心的过分依赖性;二是算法需要预先知道实际的聚类数目,而在实际应用中,聚类数目却是未知的。对此提出了一种解决方法,通过仿真实验证实了该方法的可行性与有效性。     

面向目标探测的高光谱图像层次聚类波段选择

针对高光谱图像波段间的相关件高、信息冗余大从而影响目标探测的问题,提出层次聚类波段选择(HC-BBS).首先以ROC曲线线下面积(AUC)为指标确定最佳聚类个数,然后对原始波段凝聚聚类,再在聚类后的每类波段中选择最能代表该类的波段组成最终的波段子集,保证了目标探测算子获得最佳的探测效果.对AVIRIS获取的2幅真实高光谱图像进行了实验,结果表明,HC-BBS优于另外2种波段选择方法,其选出的波段分别占据全部波段的9%和3%,目标探测算子ACE和AMF的探测率较全波段分别提高了30%和15%.     

A honeybee-mating approach for cluster analysis

Cluster analysis, which is the subject of active research in several fields, such as statistics, pattern recognition, machine learning, and data mining, is to partition a given set of data or objects into clusters. K-means is used as a popular clustering method due to its simplicity and high speed in clustering large datasets. However, K-means has two shortcomings. First, dependency on the initial state and convergence to local optima. The second is that global solutions of large problems cannot be found with reasonable amount of computation effort. In order to overcome local optima problem lots of studies done in clustering. Over the last decade, modeling the behavior of social insects, such as ants and bees, for the purpose of search and problem solving has been the context of the emerging area of swarm intelligence. Honeybees are among the most closely studied social insects. Honeybee mating may also be considered as a typical swarm-based approach to optimization, in which the search algorithm is inspired by the process of marriage in real honeybee. Neural networks algorithms are useful for clustering analysis in data mining. This study proposes a two-stage method, which first uses self-organizing feature maps (SOM) neural network to determine the number of clusters and then uses honeybee mating optimization algorithm based on K-means algorithm to find the final solution. We compared proposed algorithm with other heuristic algorithms in clustering, such as GA, SA, TS, and ACO, by implementing them on several well-known datasets. Our finding shows that the proposed algorithm works better than others. In order to further demonstration of the proposed approach’s capability, a real-world problem of an Internet bookstore market segmentation based on customer loyalty is employed.     

Data mining of spectral data

This paper describes the Spektran software system intended for automated analysis of object-attribute data tables which implements data mining algorithms based on a function of rival similarity (FRiS). The Spektran system is used to analyze a set of objects (microparticles of a substance) described by spectral characteristics. The following basic problems of data mining are solved: particle clustering by similarity of their spectra, selection of the subset of the most informative spectrum channels, identifying the classes to which particles and their mixes belong and some others.     

基于运行规律和TICC算法的风电SCADA高维时序数据聚类方法

针对大型风力发电机组高维SCADA时序数据的工况识别问题,结合风电机组运行规律和TICC算法,提出一种自动分割聚类方法。从高维的SCADA数据中选取风速、转速和桨距角等少量特定参数作为初始分割聚类对象,分析特定参数的运行规律,确定风电机组理论的运行工况。选取一段特定参数的历史数据,利用TICC算法进行离线聚类分割,获得聚类的最优特征参数。将最优特征参数作为TICC算法的输入,对新的特定参数时间序列数据进行分类。最后根据特定参数时间序列的聚类结果,对未进行分割的SCADA时序数据进行聚类处理。选取某2.5 MW双馈风电机组的SCADA时间序列数据对方法进行验证,同时将所提出的方法与FCM算法、GMM算法、K-Means算法进行对比研究。实例验证和对比研究表明,所提的聚类方法充分融合理论知识和TICC算法的优点,可高效处理高维SCADA聚类分割问题,同时保证聚类结果与理论分析结果一致性。     

Classification of vertebral column disorders and lumbar discs disease using attribute weighting algorithm with mean shift clustering

In this article, a new data pre-processing method has been suggested to detect and classify vertebral column disorders and lumbar disc diseases with a high accuracy level. The suggested pre-processing method is called the Mean Shift Clustering-Based Attribute Weighting (MSCBAW) and is based primarily on mean shift clustering algorithm finding the number of the sets automatically. In this study, we have used two different datasets including lumbar disc diseases (with two classes-our database) and vertebral column disorders datasets (with two or three classes) taken from UCI (University of California at Irvine) machine learning database to test the proposed approach. The MSCBAW method is working as follows: first of all, the centres of the sets automatically for each characteristics in dataset by using the mean shift clustering algorithm are computed. And then, the mean values of each property in dataset are calculated. The weighted datasets by multiplying these mean values by each property value in the dataset that have been obtained by dividing the above mentioned mean values by the centres of the sets belonging to the relevant property are achieved. After the data weighting stage, three different classification algorithms that included the k-NN (k-Nearest Neighbour), RBF–NN (Radial Basis Function–Neural Network) and SVM (Support Vector Machine) classifying algorithms have been used to classify the datasets. In the classification of vertebral column disorders dataset with two classes (normal or abnormal), while the obtained classification accuracies and kappa values were 78.70% ± 0.455 (the classification accuracy ± standard deviation), 81.93% ± 0.899, and 80.32% ± 0.56 using SVM, k-NN (for k = 1), and RBF–NN classifiers, respectively, the combinations of MSCBAW and SVM, k-NN (for k = 1), and RBF–NN classifiers were obtained 99.03% ± 0.977, 99.67% ± 0.992, and 99.35% ± 0.9852, respectively. In the classification of second dataset named vertebral column disorders dataset with three classes (Normal, Disk Hernia, and Spondylolisthesis), while the obtained classification accuracies and kappa values were 74.51% ± 0.581, 78.70% ± 0.659, and 83.22% ± 0.728 using SVM, k-NN (for k = 1), and RBF–NN classifiers, respectively, the combinations of MSCBAW and SVM, k-NN (for k = 1), and RBF–NN classifiers were obtained 99.35% ± 0.989, 96.77% ± 0.948, and 99.67% ± 0.994, respectively. As for the lumbar disc dataset, while the obtained classification accuracies and kappa values were 94.54% ± 0.974, 94.54% ± 0.877, and 93.45% ± 0.856 using SVM, k-NN (for k = 1), and RBF–NN classifiers, respectively, the combinations of MSCBAW and SVM, k-NN (for k = 1), and RBF–NN classifiers were obtained 100% ± 1.00, 99.63% ± 0.991, and 99.63% ± 0.991, respectively. The best hybrid models in the classification of vertebral column disorders dataset with two classes, vertebral column disorders dataset with three classes, and lumbar disc dataset were the combination of MSCBAW and k-NN classifier, the combination of MSCBAW and RBF–NN classifier, and the combination of MSCBAW and SVM classifier, respectively.