A spectral clustering algorithm for manufacturing cell formation

A graph clustering algorithm constructs groups of closely related parts and machines separately. After they are matched for the least intercell moves, a refining process runs on the initial cell formation to decrease the number of intercell moves. A simple modification of this main approach can deal with some practical constraints, such as the popular constraint of bounding the maximum number of machines in a cell. Our approach makes a big improvement in the computational time. More importantly, improvement is seen in the number of intercell moves when the computational results were compared with best known solutions from the literature.     

An e-Learning tool considering similarity measures for manufacturing cell formation

Manufacturing cell formation is the first step in the design of cellular manufacturing system. The primary objective of this step is to cluster machines into machine cells and parts into part families so that the minimum of intercell trips will be achieved. This paper will be focused on the configuration of machine cells considering three types of initial machine-part matrix: binary (zero-one) matrix, production volume matrix, and operation time matrix. The similarity measure uses only information from these types of matrix. A pure combinatorial programming formulation will be developed to maximize the sum of similarity coefficients between machine/part pairs. An e-Learning tool/application to help industrial students and engineers for enhancing their cell formation capability is proposed. This tool is designed to include a novel similarity coefficient-based heuristic algorithm for solving the cell formation problem. To determine the performance of the proposed tool, comparison is made with a well-known tool along a case study.     

基于度值和聚类系数的跨单元调度问题优化

针对工业互联网大环境下的跨单元调度存在协作效率差、生产成本过高等问题,在机器设备归置存在重叠的情况下,首先使用分层网络设计思想构造以机器和制造单元为节点的双层有向加工网络,通过分析网络中全局协作效率、单元间冗余加工路径与一阶度值的相关性,构建最小化平均度值、完工时间和加工成本的多目标调度模型.其次根据麻雀搜索算法局部搜索能力强的特点,提出了一种非支配排序遗传算法和麻雀搜索算法融合策略以及基于聚类系数的初始解生成机制.最后通过实例计算说明网络特征与跨单元调度目标呈相关性,所提模型和算法求解质量更高.     

非授权频段长期演进系统中的混合动态分簇算法

针对非授权频段长期演进（LTE）系统中动态子帧配置引起的交叉子帧干扰问题,提出了一种综合考虑大尺度损耗及小区业务量情况的混合动态分簇算法。首先,通过基站端对大尺度损耗及小区业务量情况的周期性测量,计算出对应的相关度值;然后,根据相关度值对小区进行轮询式分簇,实现小区分簇结果的周期性更新;最后,根据更新后的小区分簇结果执行动态子帧配置。仿真实验中,相比传统的静态分簇算法,中业务到达率条件下混合动态分簇算法的用户上下行平均吞吐量分别提升了约16.92%和34.33%;用户上下行平均时延分别降低了约14.18%和36.32%。仿真结果表明,混合动态分簇算法可以有效减小交叉子帧干扰的影响,提升系统吞吐量,性能优于传统的静态分簇算法。     

浅谈数据挖掘在机房信息管理中的应用

以往的机房信息管理系统,多数是OLTP系统,缺乏综合分析、辅助决策的能力,尤其对其历史积累的海量信息中隐含知识的利用无能为力。该文在聚类规则挖掘理论与研究的基础上,将k-means的算法应用于安徽经济管理学院的机房信息管理系统,对课外上机人次数据进行挖掘,根据挖掘规则得出了课外上机与课内上机时间安排的合理顺序。     

基于MapReduce和IFOA的并行密度聚类算法

针对大数据下密度聚类算法中存在的数据划分不合理、参数寻优能力不佳、并行性能较低等问题,提出一种基于IFOA的并行密度聚类算法(density-based clustering algorithm by using improve fruit fly optimization based on MapReduce,MR-DBIFOA)。首先,该算法基于KD树,提出网格划分策略(divide gird based on KD tree,KDG)来自动划分数据网格;其次在局部聚类中,提出基于自适应搜索策略(step strategy based on knowledge learn,KLSS)和聚类判定函数(clustering criterion function,CCF)的果蝇群优化算法(improve fruit fly optimization algorithm,IFOA);然后根据IFOA进行局部聚类中最优参数的动态寻优,从而使局部聚类的聚类效果得到提升;同时结合MapReduce模型提出局部聚类算法DBIFOA(density-based clustering algorithm using IFOA);最后提出了基于QR-tree的并行合并局部簇算法(cluster merging algorithm by using MapReduce,MR-QRMEC),实现局部簇的并行合并,使算法整体的并行性能得到加强。实验表明,MR-DBIFOA在大数据下的并行效率更高,且聚类效果更好。     

协作多点传输系统中基于图论的动态分簇算法

在蜂窝移动通信系统中,小区间的干扰严重限制着小区边缘用户的性能,而协作多点传输（Coordinated Multi-Point,CoMP）技术可以显著减少小区之间的干扰并改善边缘用户的性能。为了提升小区边缘用户的数据传输速率,本文针对CoMP系统提出一种基于图论的动态分簇算法。该算法利用图论的方法建立蜂窝网络的拓扑结构图,通过对小区间干扰的分析,能够同时生成多个簇大小不固定的协作簇,解决了簇大小固定和依次分簇所造成的系统受限问题。仿真结果表明,相比于其他分簇算法,本文算法在改善分簇的性能的同时降低了计算复杂度,并提高了系统的和速率。     

基于密度网格的数据流聚类算法

针对基于密度网格的数据流聚类算法中存在的缺陷进行改进,提出一种基于D-Strcam算法的改进算法NDD-Stream。算法通过统计网格单元的密度与簇的数目,动态确定网格单元的密度阂值;对位于簇边界的网格单元采用不均匀划分,以提高簇边界的聚类精度。合成与真实数据集上的实验结果表明,算法能够在数据流对象上取得良好的聚类质量。     

A grey-based clustering algorithm and its application on fuzzy system design

A grey-based clustering method was proposed and applied on fuzzy system design. A new grey-clustering algorithm using grey relational analysis as the similarity measure was developed for data clustering. It was more effective and accurate than C-Means like algorithms when dealing with data clustering issue, when the compact and complete separate data were considered. Some data clustering examples are presented to illustrate the effectiveness of the proposed clustering algorithm. Next, an application of the proposed method on fuzzy system design is presented. The procedure of fuzzy system design can be separated into two parts. In the first procedure, the grey-clustering algorithm was employed to form a rough fuzzy system only from gathered input-output data. Then, the gradient descent method was used to determine a suitable parameter set of the formed fuzzy system. A nonlinear system modelling and an inverted pendulum control problem were then used to illustrate the validity of the proposed fuzzy system design procedure.     

云计算环境下基于分形的聚类融合算法研究

传统的聚类算法不适用于处理海量和高维数据。针对云计算环境下,利用集群系统的并行计算能力,实现海量数据的聚类问题,给出了云计算环境下基于分形维数的聚类融合算法。该算法首先对基于分形维数的聚类算法进行改进,使之更适用于并行计算,其产生聚类作为初始聚类成员;再结合投票算法的融合策略实现融合。最后,对基于分形维数的聚类融合算法在云计算环境下实现并行计算。通过在UCI数据集上的对比实验来验证该算法的有效性。     

A spatiotemporal dynamic analyses approach for dockless bike-share system

The landscape of cycling activities from a dockless bike-share system is dynamic over space and time. Decoding usage patterns from bike-share trips have been a highly charged area in the literature. Therefore, this study aims at developing an analytical approach to understanding the trip demands of bike-share and model the spatiotemporal dynamics of cycling flows. Under the proposed framework, global and local Moran's I indexes measure the spatial autocorrelation of cycling trips in different traffic zones, and community detection extracts the network structure of bike traffic. The developed approach is subsequently applied to the dockless bike-share system in Singapore. It is found that the spatial distribution of the cycling trips shows significant clustering pattern. Specifically, the global Moran's indexes of weekdays are larger than that of the weekends in the same time span and, moreover, the global Moran's indexes of peak hours are smaller than the off-peak hours on the same day. Several hotspots with the top high local Moran's I values are detected, which keep relatively stable during different times of the day on both weekdays and weekends. We also found that there existed a stable community structure of bike-share trips. In contrast, the average sizes of the top 15 communities on a weekday were statistically higher than those on the weekend, at a significance level of 0.01. The proposed modeling framework provides practice insights to bike fleet management, cycling path design, and other urban and transportation planning practices.     

Integration of Artificial Immune System and K-Means Algorithm for Customer Clustering

This study proposes a novel artificial immune system (AIS)-based clustering algorithm, which integrates with a K-means (AISK) algorithm for a customer clustering problem. Computational results using Iris, Glass, Wine, and Breast Cancer benchmark datasets indicate that the proposed AIS-based clustering algorithm is more accurate than some particle swarm optimization (PSO)-based clustering algorithms. In addition, the model evaluation results using a daily transaction database provided by a cyberstore also show that the proposed AISK algorithm is superior to PSO-based clustering algorithms.     

LTE-A飞蜂窝系统干扰协调智能优化算法

在同频组网的LTE-A飞蜂窝系统中,飞蜂窝基站的密集部署会造成较为严重的同频干扰,导致网络吞吐量和用户的服务质量(Quality of Service,QoS)降低。部分频率复用(Fractional Frequency Reuse,FFR)作为常用的干扰协调方案,可以有效地提高边缘用户的服务质量。在FFR方案的基础上,通过结合遗传算法和基于模拟退火的图着色算法,提出了一种智能优化部分频率复用(Intelligence-FFR,I-FFR)算法。该算法能够动态地调整中心区域所占比例和边缘区域的频率复用因子,以增加宏小区吞吐量,降低小区边缘区域用户的中断概率。仿真结果表明,与FFR-3干扰协调算法相比,提出的I-FFR算法可使宏小区吞吐量提升15%,同时边缘区域平均用户的中断概率从85%降低到40%。     

利用模糊满意聚类建立pH 中和过程模型

利用模糊聚类方法建立 p H中和过程模型。针对模糊聚类中普遍存在的聚类个数需事先给定和收敛速度慢等问题 ,在原有聚类方法的基础上提出一种模糊满意聚类算法。该算法能快速确定系统的模糊划分数目 ,进而对应聚类个数建立相应的 TS局部线性化模型。以典型 p H中和过程为研究对象 ,利用上述方法建立其系统模型 ,取得了良好的仿真效果 ,验证了该聚类算法的快速性和有效性     

并行谱聚类算法

谱聚类算法是建立在谱图理论上的一种点对聚类算法,具有实现简单、理论基础扎实和适应任意数据空间的优点,因而成为机器学习领域的研究热点.谱聚类算法最大的问题在于计算复杂度过高,而并行计算可以提高解题效率,因此本文采用最为流行的并行计算框架MAP/REDUCE在Hadoop环境中实现了并行谱聚类算法,大大改善了谱聚类算法在大规模数据环境中的聚类效率问题.     

Development of bacteria foraging optimization algorithm for cell formation in cellular manufacturing system considering cell load variations

The cellular manufacturing system (CMS) is considered as an efficient production strategy for batch type production. The CMS relies on the principle of grouping machines into machine cells and grouping machine parts into part families on the basis of pertinent similarity measures. The bacteria foraging optimization (BFO) algorithm is a modern evolutionary computation technique derived from the social foraging behavior of Escherichia coli bacteria. Ever since Kevin M. Passino invented the BFO, one of the main challenges has been the employment of the algorithm to problem areas other than those of which the algorithm was proposed. This paper investigates the first applications of this emerging novel optimization algorithm to the cell formation (CF) problem. In addition, for this purpose matrix-based bacteria foraging optimization algorithm traced constraints handling (MBATCH) is developed. In this paper, an attempt is made to solve the cell formation problem while considering cell load variations and a number of exceptional elements. The BFO algorithm is used to create machine cells and part families. The performance of the proposed algorithm is compared with a number of algorithms that are most commonly used and reported in the corresponding scientific literature such as K-means clustering, the C-link clustering and genetic algorithm using a well-known performance measure that combined cell load variations and a number of exceptional elements. The results lie in favor of better performance of the proposed algorithm.     

A bacteria foraging algorithm based cell formation considering operation time

The cellular manufacturing system (CMS) is considered as an efficient production strategy for batch type production. The CMS relies on the principle of grouping machines into machine cells and grouping machine parts into part families based on pertinent similarity measures. The bacteria foraging algorithm (BFA) is a new in development computation technique extracted from the social foraging behavior of Escherichia coli (E. coli) bacteria. Ever since Kevin M. Passino invented the BFA, one of the main challenges has been employment of the algorithm to problem areas other than those for which the algorithm was proposed. This research work inquires the first applications of this emerging novel optimization algorithm to the cell formation (CF) problem. In addition, a newly developed BFA-based optimization algorithm for CF is discussed. In this paper, an attempt is made to solve the cell formation problem meanwhile taking into consideration number of voids in cells and a number of exceptional elements based on operational time of the parts required for processing in the machines. The BFA is suggested to create machine cells and part families. The performance of the proposed algorithm is compared with a number of algorithms that are most commonly used and reported in the corresponding scientific literature such as similarity coefficients methods (SCM), rank order clustering (ROC), ZODIAC, GRAFICS, MST, GATSP, GP, K-harmonic clustering (KHM), K-means clustering, C-link clustering, modified ART1, GA (genetic algorithm), evolutionary algorithm (EA), and simulated annealing (SA) using defined performance measures known as modified grouping efficiency and grouping efficacy. The results lie in favor of better performance of the proposed algorithm.     

A clonal selection algorithm for urban bus vehicle scheduling

The bus vehicle scheduling problem addresses the task of assigning vehicles to cover the trips in a timetable. In this paper, a clonal selection algorithm based vehicle scheduling approach is proposed to quickly generate satisfactory solutions for large-scale bus scheduling problems. Firstly, a set of vehicle blocks (consecutive trips by one bus) is generated based on the maximal wait time between any two adjacent trips. Then a subset of blocks is constructed by the clonal selection algorithm to produce an initial vehicle scheduling solution. Finally, two heuristics adjust the departure times of vehicles to further improve the solution. The proposed approach is evaluated using a real-world vehicle scheduling problem from the bus company of Nanjing, China. Experimental results show that the proposed approach can generate satisfactory scheduling solutions within 1 min.     

基于群限制的Ad Hoc网络多跳分群算法

针对建立有效的Ad Hoc网络的分群结构,达到有效使用信道、提高系统容量和网络性能的目的,在最大连通度分群算法的基础上,提出一种改进算法,该算法利用广播信道以及限定群的大小,实现了节点到群首的多跳分群。对算法进行了仿真和性能分析,结果表明,新算法保持了更加合理的分群数量,提高了群首的负载平衡性能。     

一种新型的基于密度和栅格的聚类算法*

针对网格和密度方法的聚类算法存在效率和质量问题,给出了密度和栅格相结合的聚类挖掘算法,即基于密度和栅格的聚类算法DGCA（density and grid based clustering algorithm）。该算法首先将数据空间划分为栅格单元,然后把数据存储到栅格单元中,利用DBSCAN密度聚类算法进行聚类挖掘;最后进行聚类合并和噪声点消除,并将局部聚类结果映射到全局聚类结果。实验通过人工数据样本集对该聚类算法进行理论上验证,表明了该算法在时间效率和聚类质量两方面都得到了提高。