A Probabilistic Ontology-Based Model for Multiple Aerial Target Threat Assessments in Short-Range Air Defense

Threat assessment of aerial targets is a crucial task in short-range air defense systems. When multiple targets are involved, situational information concerning entities from both sides must be effectively fused. Moreover, the dynamic changes and inherent uncertainties in the number, status, and operational intent of aerial targets further increase the challenge of assessing threats from multiple aerial targets. To obtain a more comprehensive understanding of aerial threats, a probabilistic ontology-based model for threat assessment of multiple aerial targets is developed using the Probabilistic Web Ontology Language. First, situational reasoning requirements at single-entity, dual-entity, and multientity levels are analyzed. Key situational elements in short-range air defense and their attribute, semantic, and causal relationships are then identified. Subsequently, the multientity Bayesian network (MEBN) probabilistic logic is used to model uncertain relationships among situational elements, resulting in a probabilistic ontology composed of 14 MEBN fragments, in which spatiotemporal contexts link interentity situations. Finally, a case study is conducted to evaluate the proposed model. Compared with a dynamic Bayesian network-based approach, the model provides more consistent and reasonable threat assessment results.