论文部分内容阅读
对贝叶斯网融合模型的机载多传感器性能评估进行了研究。首先对机载传感器的特性进行了分析,利用专家知识进行网络建模。给出了一种利用本地混淆矩阵(LCM)和K次测量进行计算得出的全局混淆矩阵(GCM)计算方法,随后又给出了一种利用K次GCM和LCM迭代计算K+1次GCM的计算方法。给出一些性能指标并利用迭代计算方法对模型分别进行了目标属性未知、目标属性为我、目标属性为敌三种场景下各传感器性能计算,对结果进行了分析并和真实值进行了比较分析,证明了结果的准确性。
The airborne multi-sensor performance evaluation of Bayesian network fusion model was studied. First of all, the characteristics of airborne sensors were analyzed, using the expert knowledge to network modeling. A global confusion matrix (GCM) calculation method using locally confused matrices (LCMs) and K measurements is given, followed by an iterative calculation of K + 1 GCMs using K GCMs and LCMs The calculation method. Some performance indexes are given and the performance of each model is evaluated separately based on the performance of the sensor with unknown target property, target property, and target property. The results are analyzed and compared with the real values , Proved the accuracy of the result.