中文摘要：

针对传统贝叶斯网络对复杂对象建模存在的问题,分析了面向对象的概率关系模型。建立基于面向对象概率关系模型知识库的贝叶斯网络生成算法,建立战场威胁级别评估模型,并实现对威胁级别的评估。仿真结果证明了该方法的有效性,并且具有比贝叶斯网络模型更强的建模能力。

英文摘要：

Considering some of the limitations of Bayesian Network（BN） in modeling complex domains,the Object-Oriented Probabilistic Relational Model（OPRM） was analyzed.Then an algorithm for generating Bayesian Networks from the knowledge base of OPRM was proposed.A model for threat level assessment was developed and the result of threat level assessment was given.Simulation results demonstrate that the proposed method is effective and more powerful in modeling complex objects than Bayesian Networks.