To address the current problems of poor generality,low real-time,and imperfect information transmission of the battlefield target intelligence system,this paper studies the battlefield target intelligence system from ...To address the current problems of poor generality,low real-time,and imperfect information transmission of the battlefield target intelligence system,this paper studies the battlefield target intelligence system from the top-level perspective of multi-service joint warfare.First,an overall planning and analysis method of architecture modeling is proposed with the idea of a bionic analogy for battlefield target intelligence system architecture modeling,which reduces the difficulty of the planning and design process.The method introduces the Department of Defense architecture framework(DoDAF)modeling method,the multi-living agent(MLA)theory modeling method,and other combinations for planning and modeling.A set of rapid planning methods that can be applied to model the architecture of various types of complex systems is formed.Further,the liveness analysis of the battlefield target intelligence system is carried out,and the problems of the existing system are presented from several aspects.And the technical prediction of the development and construction is given,which provides directional ideas for the subsequent research and development of the battlefield target intelligence system.In the end,the proposed architecture model of the battlefield target intelligence system is simulated and verified by applying the colored Petri nets(CPN)simulation software.The analysis demonstrates the reasonable integrity of its logic.展开更多
针对海战场环境下态势评估中目标数量多、类型复杂多样的问题,首先引入数据聚类对态势评估的目标分群环节进行聚类分群,提出基于DBSCAN(density-based spatial clustering of applications with noise)算法的密度聚类,可聚类任意形状的...针对海战场环境下态势评估中目标数量多、类型复杂多样的问题,首先引入数据聚类对态势评估的目标分群环节进行聚类分群,提出基于DBSCAN(density-based spatial clustering of applications with noise)算法的密度聚类,可聚类任意形状的数据簇,遍历性好,能够对战场环境下目标进行全面合理的分群;然后,给出了算法计算的基本步骤,并利用算例对已知战场态势的目标群进行正确性验证;最后,将该算法与基于划分的K-means算法、基于层次的AGNES(AGglomerative NESting)算法进行了对比分析,证明了该算法的有效性和合理性。展开更多
由于海面目标所处的海战场环境的复杂性,以及海面目标探测获取的状态数据的不确定、缺失、模糊以及动态变化等,使得海面目标的综合识别非常困难。为了解决这些问题,需要对海战场多传感器目标综合识别中采用智能的神经网络进行应用研究...由于海面目标所处的海战场环境的复杂性,以及海面目标探测获取的状态数据的不确定、缺失、模糊以及动态变化等,使得海面目标的综合识别非常困难。为了解决这些问题,需要对海战场多传感器目标综合识别中采用智能的神经网络进行应用研究。文中提出一种特征级采用LMBP神经网络算法,从多维度学习训练多传感器获取的数据,提取更多相关细节的目标特征属性信息,决策级采用证据理论的智能算法模型,即列文伯格误差反向传播结合证据理论算法的目标综合识别算法(Levenberg Marquardt Back Propagation-Dempster Shafer,简称LMBP-DS算法);然后通过Matlab仿真实验,比较不同的隐藏节点数对识别率的影响,找到LMBP-DS算法最佳的神经网络结构;通过对比实验得出:LMBP-DS算法比动量自适应学习BP-DS算法具有更快的收敛速度,同时具有更稳健高效的识别正确率,从而更适用于海战场多传感器目标综合识别。展开更多
基金supported by the National Natural Science Foundation of China(41927801).
文摘To address the current problems of poor generality,low real-time,and imperfect information transmission of the battlefield target intelligence system,this paper studies the battlefield target intelligence system from the top-level perspective of multi-service joint warfare.First,an overall planning and analysis method of architecture modeling is proposed with the idea of a bionic analogy for battlefield target intelligence system architecture modeling,which reduces the difficulty of the planning and design process.The method introduces the Department of Defense architecture framework(DoDAF)modeling method,the multi-living agent(MLA)theory modeling method,and other combinations for planning and modeling.A set of rapid planning methods that can be applied to model the architecture of various types of complex systems is formed.Further,the liveness analysis of the battlefield target intelligence system is carried out,and the problems of the existing system are presented from several aspects.And the technical prediction of the development and construction is given,which provides directional ideas for the subsequent research and development of the battlefield target intelligence system.In the end,the proposed architecture model of the battlefield target intelligence system is simulated and verified by applying the colored Petri nets(CPN)simulation software.The analysis demonstrates the reasonable integrity of its logic.
文摘针对海战场环境下态势评估中目标数量多、类型复杂多样的问题,首先引入数据聚类对态势评估的目标分群环节进行聚类分群,提出基于DBSCAN(density-based spatial clustering of applications with noise)算法的密度聚类,可聚类任意形状的数据簇,遍历性好,能够对战场环境下目标进行全面合理的分群;然后,给出了算法计算的基本步骤,并利用算例对已知战场态势的目标群进行正确性验证;最后,将该算法与基于划分的K-means算法、基于层次的AGNES(AGglomerative NESting)算法进行了对比分析,证明了该算法的有效性和合理性。
文摘由于海面目标所处的海战场环境的复杂性,以及海面目标探测获取的状态数据的不确定、缺失、模糊以及动态变化等,使得海面目标的综合识别非常困难。为了解决这些问题,需要对海战场多传感器目标综合识别中采用智能的神经网络进行应用研究。文中提出一种特征级采用LMBP神经网络算法,从多维度学习训练多传感器获取的数据,提取更多相关细节的目标特征属性信息,决策级采用证据理论的智能算法模型,即列文伯格误差反向传播结合证据理论算法的目标综合识别算法(Levenberg Marquardt Back Propagation-Dempster Shafer,简称LMBP-DS算法);然后通过Matlab仿真实验,比较不同的隐藏节点数对识别率的影响,找到LMBP-DS算法最佳的神经网络结构;通过对比实验得出:LMBP-DS算法比动量自适应学习BP-DS算法具有更快的收敛速度,同时具有更稳健高效的识别正确率,从而更适用于海战场多传感器目标综合识别。
