基于置信规则库的防空目标意图识别方法

为解决防空反导体系对于目标意图识别中多源信息处理的问题,提出了一种新的防空目标识别方法——置信规则库(Belief Rules Base)识别模型,该方法是基于BRB专家知识系统的自学习方法,可以有效处理多源不确定性信息。同时,为了提高其识别效率,建立了一种与之相匹配的参数优化模型,并选择差分进化算法作为BRB识别模型的优化引擎。此外,在BRB识别模型中,专家意见也被引用到模型的初始化和推理过程中。最后采用一个实例对BRB识别模型的准确度进行验证。研究表明:该方法具有较强的实用性,可为防空反导体系的目标意图识别提供有力支撑。

Intension Identification in Air Defense Based on Belief Rule Base Expert System under Expert Guidance

To solve the fusion of multi-source information in intension identification, a new identifier is proposed based on the Belief Rule Base (BRB) expert system and the evolutionary algorithms, named as BRB-identifier. To improve its efficiency, a corresponding BRB-identifier parameter optimization model is constructed. The Differential Evolutionary(DE)algorithm are tested as the optimization engine to solve the optimization model. Moreover,expert guidance is introduced in both the BRB-identifier initialization and inference process. An example demonstrates that the model and algorithm is reasonable in the end. The efficiency of the proposed BRB-identifier is validated by a practical intention identification of attacking suspected object case in the end. Research shows that the BRB-identifier can support for intension identification in Air Defense.