【作者】 周倜；

【导师】 王小非；

【作者基本信息】 哈尔滨工程大学 ， 计算机应用技术， 2013， 博士

【摘要】 随着电磁技术和用频设备的广泛应用，战场电磁环境变得日益复杂、电磁频谱资源越来越重要。特别是在信息化元素极为丰富的海战场，复杂电磁环境给海上作战指挥提出了更高的挑战，这在很大程度上体现在对电磁环境认知和全面电磁态势的需求。作为传统战场态势的重要补充，电磁态势覆盖了从辐射源信息融合到频谱资源调配、从战术级方案优化到战役级辅助决策等作战的各个方面，准确认识海战场电磁环境，并在态势理解和态势展现的基础上生成全面的电磁态势，是夺取海上制电磁权的关键因素。本文从海战场指挥控制和频谱管理对电磁态势的实际需求出发，研究了海战场电磁态势生成的体系结构，提出了一种海战场电磁态势生成框架，研究了该框架下一些重要过程域的关键技术，包括电磁环境展现技术、电磁态势分析技术和电磁态势评估技术，并给出了仿真实验结果及相关理论分析。本文的主要贡献和创新点具体体现在以下几个方面：（1）建立了一种海战场电磁态势生成框架。基于传统态势生成模型中对数据分级处理的原则，提出了一套包含电磁环境感知、电磁环境构建、电磁态势理解和电磁态势展现四个过程域的电磁态势生成框架。其中，定义了战场电磁态势的相关概念，给出了面向电磁环境构建的用频装备模型、电磁环境计算方法和电磁环境描述方法，介绍了电磁态势理解域中态势分析和态势评估的概念、目标、及基本技术路线。统一的电磁态势生成框架利于各关键技术的有机结合、态势生成的完整性和规范性。（2）提出了一种时间自适应的多维电磁环境态势展现方法。将区域电磁环境转化为由离散像素点组成的位图，从而构建典型频谱特征直方图，通过计算直方图特征向量距离来度量电磁环境的变化程度。基于对电磁环境动态性的度量，利用平行坐标系方法实现了电磁环境态势的多维展现和时间自适应性。对传统的平行坐标系方法进行了改进，包括数据项拟合、定量映射、自适应时间滑窗三个部分的算法理论模型。其中基于Bezier曲线的数据拟合用以解决各维度间数据项连接时的数据遮蔽混淆以及电磁态势各个维度之间信息关联不清晰的问题；定量映射策略模型将电磁态势的多个维度的数据项定量映射到平行坐标系中的相应坐标轴上；自适应步长的时间滑窗机制能够有效调整电磁态势展现的时间间隔，使其充分适应电磁环境的动态变化。实现了多维电磁环境态势在单幅图表中的快速、高效展现。（3）提出了一种面向大量用频节点的电磁干扰关系网络建模方法。针对海战场舰艇编队中用频装备众多的特点以及编队频管的实际需求，提出了基于复杂网络的电磁干扰关系建模思想，并建立了两种干扰关系网络模型：基于实时解析法的干扰关系网络（A型EMIN）、基于经验统计法的干扰关系网络（B型EMIN）。通过对两种EMIN的仿真及网络特性分析，指出了它们各自特点以及在电磁干扰态势分析中的应用方向。通过对B型EMIN聚类性质的仿真分析，验证了分层频管的合理性。基于平均距离、度、聚集系数和网络密度等网络特征提出了相应的EMIN度量指标，解释了这些特征和指标在EMIN中的具体含义，并指出了它们在评价和度量EMIN中的有效性。（4）提出了一种海战场电磁态势评估模型。该模型包括一套评估指标和两种评估方法。综合考虑影响海战场电磁域作战的各个因素，定义了一套电磁态势评估指标。针对海战场编队的结构特点，提出了一种层次化的指标定量评估方法，通过对具体装备的测量和计算，融入了装备和任务权重的计算结果逐层向上汇聚，形成该指标在编队层面的综合评价，并以编队干扰程度评估的仿真实例验证了方法的有效性。针对电磁态势中各指标相互影响的特点，提出了一种基于模糊认知图（FCM）的综合态势评估方法，给出了FCM邻接矩阵的量化方法，改进了传统FCM的阈值函数和推理方法使其适用于电磁态势评估。相比传统电磁环境评估手段，FCM模型考虑了更全面的态势指标以及指标间的关联关系，评估结果更加准确，且具备一定的态势预测能力。更多还原

【Abstract】 With the widespread use of electromagnetic technology and electronic equipment,electromagnetic environment (EME) of battlefield has become increasingly complex, andspectrum has become more and more important. Especially in the naval battlefield with a highlevel of information technology, EME has put forward higher requests to command andcontrol—that is, to a considerable degree, demands for comprehensive electromagneticsituation (EMS). As an important complement to conventional battlefield situation, EMSplays a key role in various battle domains, such as emitter data fusion, spectrum deployment,schema optimization in tactic-level, decision supporting in campaign-level, and so on.Therefore, electromagnetic situation generation (EMSG), consisting of environmentunderstanding, situational awareness, and situation representation, has become a critical factorin electromagnetic supremacy.According to the needs of EMS in naval Command&Control and spectrummanagement (SM), this thesis studies the architecture of EMSG, and proposes a frameworkfor EMSG in sea battlefield. Based on deeply research on several key techniques in EMSG,including situation representation, situation analysis, and situation assessment, we proposesome noval models and methods, and also present the theoretical analysis and the experimentresults. The main innovative achievements are as follows:(1) An electromagnetic situation generation framework (EMSGF) for sea battlefield isbuilt. Based on hierarchical processing principle in traditional situation models, the EMSGFconsists of four process domains—electromagnetic environment awareness, electromagneticenvironment constructing, electromagnetic situation understanding, and electromagneticsituation representation. In addition, we define the concepts and terms of EMS and its processdomains. Models for typical equipment and propagation loss, and an EM characterizationmethod are presented to construct the virtual electromagnetic environment. We also introducethe concepts, objectives, and technology of situation analysis and situation assessment, whichare the essential components of situation understanding. The proposed EMSGF shouldcontribute to the integrity and normalization of EMSG.(2) A time-adaptive visualization approach for EME situation representation is proposed.Expressing a specified electromagnetic area in the form of bitmap consisting of pixels, we cancreate the histogram of spectrum features. Then the change in frequency-domain of thespecified EM area can be measured by calculating the distance of feature vectors between histograms. Based on the measurement of variability of EME and parallel coordinates method,a time-adaptive visualization approach for multi-dimensional EME situation representation isrealized. We propose an Advanced Parallel Coordinates Method (APCM) to build avisualization framework for multi-dimensional EMS representation. It mainly includes threealgorithmic modules of a data fitting, a quantitative mapping, an adaptive time-slidingwindow. Respectively, the data fitting model mitigates the degree of line overlapping betweenthe correlated adjacent data dimensions by Bezier curve. The quantitative mapping model canmap the multi-dimensional information in different sizes and types into coordinate system.Based on them, the adaptive time-sliding window model can adjust time interval of situationrepresentation dynamically in order to make it adaptable to the change in an electromagneticenvironment. The experiments show that the APCM is an effective approach to displaymulti-dimensional EMS information quickly and precisely.(3) For large number of spectrum-depending nodes, this thesis proposes a modelingapproach for relation network of electromagnetic interference. In consideration of the largenumber of electronic equipment in naval fleet, we build a complex network model, in whichelectronic equipment are nodes and interference relation are edges. Two forms of the modelare presented—one with real-time analyze, the other with experiential statistic. The twomodels are named as A-EMIN and B-EMIN, respectively. A series of simulations and analysisof the two EMINs reveal some useful network characteristic that make sense to bothinterference analysis and spectrum management. The simulation result of cluster characteristicof B-EMIN verifies the rationality of a hierarchical SM strategy. In addition, we presentseveral EMIN metrics based on network features such as degree, density, mean distance,clustering coefficient, and so on. Explaining what these metrics mean in EMINs, we point tothe application potential of them in EMIN evaluation.(4) An electromagnetic situation assessment system for sea battlefield is set up. Takinginto account of all the factors considered in EMS, we propose a relatively comprehensiveindex system for EMS assessment. According to the structural feature of naval fleet, ahierarchical, quantitative assessment model for some indices is proposed, in which thesituation data is collected from equipment-level, and flows to fleet-level in a bottom-up way.Consequently, the measure of this index on fleet-level is generated with consideration overequipment weight and task weight. Applied to evaluating the index of EMI degree in fleet, themodel is proven to be effective. An FCM based situation assessment model is proposed inconsideration of the feature that most indices influence each other. We present the quantification method for adjacency matrix of FCM, and improve the threshold function andinference algorithm to make the model applicable to EMS assessment. Compared with theconventional EME assessment approaches, the FCM takes into consideration of variousfactors and the relations between them, produces more accurate results, and has a certainability of situation prediction.更多还原