摘要
通过分析军事通信网络效能评估数据源的特征,提出了一种融合专家经验、仿真和实测数据评估军事通信网络效能的方法。包括用信念图统一表达数据形式;以实测数据为训练样本,使用支持向量机拟合网络参数与性能指标关系模型,通过比较预测值与仿真结果获取仿真数据可信度;采用扩展贝叶斯法融合多源数据,给出了基于多数据源的军事通信网络效能评估流程;最后以评估某舰艇编队军事通信网络效能为例,给出了具体融合过程,结果表明融合多数据源的评估结果的有效性和可信性更高。
A novel method which synthesizes the expert experiences, simulation and test bed data to evaluate the effectiveness of military communication network with respect to the characteristics of multiple data sources are proposed. Using belief map as data description form; taking test bed data as training example; modeling regression relationship of network parameters and performance measures and acquiring simulation data^s beliefs based on SVR; using extensive Bayesian algorithm to fuse data from multiple sources;this paper constructs an effectiveness evaluation framework of military communication network with multiple data sources. Taking military communication network effectiveness evaluation of a Naval Vessels Fleet communication as an example ,this paper demonstrates that the fusion processing proposed is credible and effective.
出处
《火力与指挥控制》
CSCD
北大核心
2012年第7期18-23,共6页
Fire Control & Command Control
基金
国家自然科学基金资助项目(61074107
91024015
60974073
60974074)
关键词
评估
融合
信念图
支持向量机
扩展贝叶斯
evaluation, fusion, belief map, support vector regression,extensive bayesian
