Target Damage Level Assessment Based On Dynamic Bayesian Networks

The target battle damage assessment (BDA) is an important part of command automation System. The complication, dynamic and uncertainty of information warfare have brought the huge challenge for the target BDA. In this case, how to carry on a comprehensive and reasonable real-time assessment to target battle damage is an urgent need to solve. To evaluate the target damage rank effectively needs a great deal of indefinite target characteristic information. How to fuse these indefinite information to realize real-time assessment to target battle damage is the key technology difficulty.In view of the advantage of the Bayesian network, particularly dynamic Bayesian network in processing indefinite information, the paper emphatically studied the target damage rank assessment method and the application which based on dynamic Bayesian network. The prime task and the content in this article can be concluded as follows:(1) The basic framework of information fusion based on dynamic Bayesian networks was discussed. With the aid of this framework, the rationality and feasibility of dynamic Bayesian network in processing indefinite information was proved by an application example.(2) The dynamic Bayesian network model of target damage rank assessment was analyzed and researched. Firstly, the target damage rank assessment was divided into the target vulnerability assessment and the target suffering attack degree assessment two processes. Then the correlated variables which influence target damage rank assessment were analyzed. In this foundation, the dynamic Bayesian network model of target damage rank assessment was established.(3) A correlated example of target damage rank assessment based on dynamic Bayesian network was analyzed, which takes the target BDA in the application of command decision as a background. The evaluation results are demonstrated the approach mentioned above feasible and reasonable.(4) The contrastive analysis was carried out, which about the accuracy as well as the fault-tolerant ability in the target damage rank assessment between Dynamic Bayesian network and static Bayesian network. The superiority of dynamic Bayesian network was proved.