态势估计中的目标分群算法设计与实现

针对态势估计中的目标分群问题,分析了目标的位置、运动状态和属性对分群结果的影响,建立了目标相似度计算模型,并提出了分群策略,包括群的形成、分裂和合并等。采用面向对象软件设计思想建立了分群对象模型,分析了对象间关系。给出了分群算法的主体流程,并对其中的关键步骤进行了详细的说明。开发了目标分群算法模块,实现了对数据的预处理、目标群的更新维护、群属性的计算等功能,通过仿真试验,调整目标分群影响因子及加权因子,得到了较好的分群效果,验证了文中方法的可行性和正确性。     

基于相似度矩阵的雷达探测目标分群算法

目前对于雷达探测的目标分群问题,在技术实现上仍然存在操作效率低、分群正确性得不到保证、适用性不强等不足,对此,本文提出了一种基于相似度矩阵的目标分群算法,该算法具有过程易于实现、分群效率高、正确率高、适用性强等优点,算法根据雷达探测目标的分布特点,定义了目标属性间的相似度计算方法,构建相似度矩阵,通过设计行列变换准则和判定准则来实现矩阵目标分群,并设计了维护规则来实现在群结构发生变化的情况下对编群结构进行动态维护,最后在此基础上,通过一个典型实例采用SPSS数学工具验证了算法的正确性、有效性与适用性。     

态势估计中目标分群方法的研究

对态势估计中目标分群问题及其相关概念进行了描述;定义了距离因子和距离差因子,根据初级融合获得的敌目标有用数据,计算进攻关系隶属度以生成敌方功能群;提出了目标分群的算法步骤,并给出了算例。仿真结果表明,该算法能够为态势估计生成期望的功能群,具有一定的理论和实用价值。     

通信时延下无人机群集系统分群控制算法

针对存在通信时延的无人机群集系统分群控制问题,设计了基于蚁群算法的协同分群控制算法。借鉴蚁群算法概率公式,利用个体自身位置信息和存在通信时延的任务信息(期望子群规模和群目标运动信息),设计了子群数量和规模可控、结构和速度调整较小的分群策略;将存在通信时延的群目标运动信息融入协同分群控制律中,实现了子群速度动态可控的分群,并利用Lyapunov稳定性定理和LaSalle不变性原理进行了稳定性分析;通过仿真实验进行了方法的有效性验证。结果表明,在通信时延约束下,所设计算法能够使群集系统实现结构和速度调整较小的可控分群行为。     

基于模糊ART划分的目标分群算法

针对目标分群中存在的分类数未知和噪声干扰问题,提出一种基于模糊ART划分的目标分群算法。通过目标识别属性划分,约减分群目标数规模,降低计算量;通过划分数据预处理消除尺度差异,在此基础上采用基于模糊ART的目标空间划分,经类选择、匹配度检验和类学习等步骤实现对目标的增量式动态分群。试验结果表明,该算法对复杂环境下未知分类数的多目标编队分群具有良好的有效性、稳健性和实时性。     

基于预测度量值的IPTV用户行为规则预测算法

提出了一种符合用户行为的,基于海量IPTV用户特征数据,对IPTV用户进行分群和规则提取的算法模型。首先提出了符合用户点播使用行为的IPTV用户分群的描述维度,即通过基础属性描述用户分群、通过点播行为描述用户分群变化趋势。然后提出了预测度量值的概念,对用户分群的稳定性进行描述,并提出了对稳定的用户分群提取点播行为概率的算法。最后通过大量的IPTV运营数据对算法模型进行了验证分析。     

基于目标分群与模糊匹配的态势评估技术

采用目标分群和部队编制模糊匹配技术实现了态势评估系统。介绍了数据融合发展概况与融合模型,在数据融合修正模型的基础上,提出了态势评估总体技术框架、功能模块和关键技术。介绍了目标分群处理流程,包括目标分群、群的分裂与合并,并进一步阐述了目标分群算法与模糊匹配算法。介绍了基于模糊匹配技术实现军事体系单元假设推理的方法,给出目标分群计算结果,说明了算法的有效性。     

态势评估中的目标编群问题研究

目标编群是态势评估中需要解决的一个重要问题。首先建立了一种相似性测度模型,该模型通过对目标平台属性进行"乘运算"和"加运算"两方面的相似测度,因而对目标间的相似性能做出快速的判别,而且能更准确有效地判断目标对象间的相似性程度,加强目标辨别能力,降低错误编群的概率。在此基础上,给出了群递增形成算法,最后以一组数据验证了该算法可快速有效实现目标编群。     

一种电磁传感器网络的群规模约束分群算法

电磁传感器网络是一种以电磁信号为感知目标的传感器网络。本文针对电磁传感网监测性能要求,提出了一种基于分群的电磁传感器网络频域协作工作模式,理论上分析了群规模与信号感知概率以及感知时延的关系,将电磁传感网监测性能要求转化为对分群规模的约束,并设计了群规模约束分群算法(CRCA)。仿真结果表明,在相同的性能要求和网络连通度下,CRCA得到的满足群规模约束的群比例高出已有算法10%,能较好地满足系统性能要求。     

一种无线传感器网络全连通群的休眠调度算法

该文针对无线传感器网络的覆盖性和连通性问题,在假设传感器节点地理位置信息已知的条件下,设计了一种包含全连通群的建立和维护以及群内节点休眠调度的全新算法。该算法采用保证群内节点彼此一跳可达的全连通群分群方法,以及分布式节能的休眠调度策略,最大程度上减少传感器网络的能量消耗,延长了网络寿命。仿真结果表明:该算法能较好地保证无线传感器网络的覆盖性和连通性,且能耗较低。     

数据流上基于K-median聚类的算法研究

文章研究和分析了数据流上的K-median聚类算法技术,包括:(1)流模型和K-median问题定义;(2)基于流的K-median聚类基本决策和内在机理;(3)理论上有性能保证的流算法。对于每一特征,这种技术能在没有实际保留任何数据流对象的情形下有效地确定聚类点。它通过一个聚类块的一分为二或相邻聚类块的合二为一来动态地生成聚类点,从而实现上述目标。作为结果,这种技术所确定的聚类点将比其他常规方法更准确。在数据流环境中,这种技术能够在产生高质量聚类结果的同时非常有效地执行。     

一种基于动态遗传算法的聚类新方法

 如何确定聚类数目一直是聚类分析中的难点问题.为此本文提出了一种基于动态遗传算法的聚类新方法,该方法采用最大属性值范围划分法克服划分聚类算法对初始值的敏感性,并运用两阶段的动态选择和变异策略,使选择概率和变异率跟随种群的聚类数目一致性变化,先进行不同聚类数目的并行搜索,再获取最优的聚类中心.七组数据聚类实验证明该方法能够实现数据集最佳划分的自动全局搜索,同时搜索到最佳聚类数目和最佳聚类中心.     

Evolution-Based Tabu Search Approach to Automatic Clustering

Traditional clustering algorithms (e.g., the K-means algorithm and its variants) are used only for a fixed number of clusters. However, in many clustering applications, the actual number of clusters is unknown beforehand. The general solution to this type of a clustering problem is that one selects or defines a cluster validity index and performs a traditional clustering algorithm for all possible numbers of clusters in sequence to find the clustering with the best cluster validity. This is tedious and time-consuming work. To easily and effectively determine the optimal number of clusters and, at the same time, construct the clusters with good validity, we propose a framework of automatic clustering algorithms (called ETSAs) that do not require users to give each possible value of required parameters (including the number of clusters). ETSAs treat the number of clusters as a variable, and evolve it to an optimal number. Through experiments conducted on nine test data sets, we compared the ETSA with five traditional clustering algorithms. We demonstrate the superiority of the ETSA in finding the correct number of clusters while constructing clusters with good validity.     

基于协同度的基站群利益树动态分簇算法

该文针对协同基站群分簇算法缺乏通用模型的问题,提出了一种协同度分簇模型,将系统和容量最大化简化为协同度最大化。在该模型的指导下,将分簇问题建模为有向带权连通图的利益树生成问题,设计了一种利益树动态分簇算法。该算法能够并行生成多个规模动态变化的协同簇,克服了传统顺序分簇导致的系统性能受限的问题;且分簇结果的协同度之和最大,可获得近似最优的分簇性能。仿真结果表明,该算法与传统贪婪搜索算法相比,系统频谱利用率提高了约0.4 bit/Hz,且算法复杂度只与基站个数呈线性关系。     

Clustering Methods for Agent Distribution Optimization

Multiagent systems consist of a collection of agents that directly interact usually via a form of message passing. Information about these interactions can be analyzed in an online or offline way to identify clusters of agents that are related. The first part of this paper is dedicated to a formal definition of a proposed dynamic model for agent clustering and experimental results that demonstrate applicability of this novel approach. The main contribution is the ability to discover and visualize communication neighborhoods of agents at runtime, which is a novel approach not attempted so far. The second part of this paper deals with a static agent clustering problem where equally sized clusters with maximal intracluster communication among agents are sought in order to efficiently distribute agents across multiple execution units. The weakness of standard clustering approaches for solving this type of clustering problem is shown. First, these algorithms optimize the generated clustering with respect to just one criterion, and therefore, yield solutions with inferior quality relative to the other criteria. Second, the algorithms are deterministic; thus they can produce just a single solution for the given data. A multiobjective clustering approach based on an iterative optimization evolutionary algorithm called multiobjective prototype optimization with evolved improvement steps (mPOEMS) is proposed and its advantages are demonstrated. The most important observation is that mPOEMS produces numerous high-quality solutions in a single run from which a user can choose the best one. The best solutions found by mPOEMS are significantly better than the solutions generated by the compared clustering algorithms.      

The rise and fall of spatio-temporal clusters in mobile ad hoc networks

Cluster detection has been widely applied to the problem of efficient data delivery in highly dynamic mobile ad hoc networks. By grouping participants who meet most often into clusters, hierarchical structures in the network are formed which can be used to efficiently transfer data between the participants. However, data delivery algorithms which rely on clusters can be inefficient in some situations. In the case of dynamic networks formed by encounters between humans, sometimes called Pocket Switched Networks (PSNs), cluster based data delivery methods may see a drop in efficiency if obsolete cluster membership persists despite changes to behavioural patterns. Our work aims to improve the relevance of clusters to particular time frames, and thus improve the performance of cluster based data delivery algorithms in PSNs. Furthermore, we will show that by detecting spatio-temporal clusters in PSNs, we can now improve on the data delivery success rates and efficiency of data delivery algorithms which do not use clustering; something which has been difficult to demonstrate in the past.     

基于人工免疫网络的动态聚类算法

聚类分析的两个基本任务是分析数据集中簇的数量以及这些簇的位置.大多数的聚类方法通常只关注后一个问题.为了在聚类数不确定的情况下实现聚类分析,本文提出了一种新的结合人工免疫网络和遗传算法的动态聚类算法—DCBIG.新算法主要包含两个阶段:先使用人工免疫网络算法获得聚类可行解,然后使用遗传算法依据聚类可行解实现动态聚类.本文对获得聚类可行解的条件和概率进行了分析.仿真实验结果表明与现有方法相比,新方法具有更高的收敛概率和收敛速度.     

Cluster‐based mobility management algorithms for wireless mesh networks

Wireless mesh networks (WMNs) have been the recent advancements and attracting more academicians and industrialists for their seamless connectivity to the internet. Radio resource is one among the prime resources in wireless networks, which is expected to use in an efficient way especially when the mobile nodes are on move. However, providing guaranteed quality of service to the mobile nodes in the network is a challenging issue. To accomplish this, we propose 2 clustering algorithms, namely, static clustering algorithm for WMNs and dynamic clustering algorithm for WMNs. In these algorithms, we propose a new weight‐based cluster head and cluster member selection process for the formation of clusters. The weight of the nodes in WMN is computed considering the parameters include the bandwidth of the node, the degree of node connectivity, and node cooperation factor. Further, we also propose enhanced quality of service enabled routing protocol for WMNs considering the delay, bandwidth, hopcount, and expected transmission count are the routing metrics. The performance of the proposed clustering algorithms and routing protocol are analyzed, and results show high throughput, high packet delivery ratio, and low communication cost compared with the existing baseline mobility management algorithms and routing protocols.     

用于数据挖掘的聚类算法

数据挖掘用于从超大规模数据库中提取感兴趣的信息。聚类是数据挖掘的重要工具,根据数据间的相似性将数据库分成多个类,每类中数据应尽可能相似。从机器学习的观点来看,类相当于隐藏模式,寻找类是无监督学习过程。目前已有应用于统计、模式识别、机器学习等不同领域的几十种聚类算法。该文对数据挖掘中的聚类算法进行了归纳和分类,总结了7类算法并分析了其性能特点。     

LIDAR: a protocol for stable and energy-efficient clustering of ad-hoc multihop networks

Clustering has been proposed as a promising method for simplifying the routing process in mobile ad hoc networks (MANETs). The main objective in clustering is to identify suitable node representatives, i.e. cluster heads (CHs) to store routing and topology information; CHs should be elected so as to maximize clusters stability, that is to prevent frequent cluster re-structuring. Since CHs are engaged on packet forwarding they are prone to rapidly drop their energy supplies, hence, another important objective of clustering is to prevent such node failures. Recently proposed clustering algorithms either suggest CH election based on node IDs (nodes with locally lowest ID value become CHs) or take into account additional metrics (such as energy and mobility) and optimize initial clustering. Yet, the former method is biased against nodes with low IDs (which are likely to serve as CHs for long periods and therefore run the risk of rapid battery exhaustion). Similarly, in the latter method, in many situations (e.g. in relatively static topologies) re-clustering procedure is hardly ever invoked; hence initially elected CHs soon suffer from energy drainage. Herein, we propose LIDAR, a novel clustering method which represents a major improvement over alternative clustering algorithms: node IDs are periodically re-assigned so that nodes with low mobility rate and high energy capacity are assigned low ID values and, therefore, are likely to serve as CHs. Therefore, LIDAR achieves stable cluster formations and balanced distribution of energy consumption over mobile nodes. Our protocol also greatly reduces control traffic volume of existing algorithms during clustering maintenance phase, while not risking the energy availability of CHs. Simulation results demonstrate the efficiency, scalability and stability of our protocol against alternative approaches.