Analyzing air combat simulation results with dynamic bayesian networks

Abstract

In this paper, air combat simulation data is reconstructed into a dynamic Bayesian network. It gives a compact probabilistic model that describes the progress of air combat and allows efficient computing for study of different courses of the combat. This capability is used in what-if type analysis that investigates the effect of different air combat situations on the air combat evolution and outcome. The utilization of the dynamic Bayesian network is illustrated by analyzing simulation results produced with a discrete event air combat simulation model called X-Brawler.

Influence diagrams in analysis of discrete event simulation data

“…Poropudas and Virtanen (2007) presents a method for analyzing the evolution of discrete event simulations with dynamic Bayesian networks (DBNs, e.g., Neapolitan 2004, Jensen and Nielsen 2007, Pearl 1986). In the presented approach, chance nodes of a DBN and their interconnecting arcs are used to describe the joint probability distribution of the random variables that presents the evolution of the simulation state during simulation.…”

“…In the presented approach, chance nodes of a DBN and their interconnecting arcs are used to describe the joint probability distribution of the random variables that presents the evolution of the simulation state during simulation. The analysis of such networks has been found more effective and less time consuming than the brute force analysis of raw simulation data (Poropudas and Virtanen 2007). Unfortunately, DBNs are only descriptive and offer no direct way for optimizing the simulation output.…”

“…The IDs are probabilistic models that are used for structuring and solving decision problems under uncertainty (e.g., Diehl andHaimes 2004, Virtanen, Raivio, andHämäläinen 1999). In the construction of an ID, a DBN is first built by combining simulation data with expert knowledge (Neapolitan 2004, Poropudas andVirtanen 2007) to present the evolution of the simulation state. Then, the DBN is extended into an ID by including the simulation parameters into the model to represent the parameters' effect on the simulation output and the changes in the simulation state.…”

“…The time evolution of the simulation state and interdependence between the states at different times are modeled with the conditional probability tables of the chance nodes. This presentation for the simulation state is in fact a DBN (Poropudas and Virtanen 2007).…”

Study on air combat tactics decision-making based on bayesian networks

“…But, the applied decision-making system also under studying, the key technology consist decision-making theory, data base, collateral computing technique and battlefield information fusion. Bayesian network as an effective method to make uncertainty reasoning in air combat has been attended widely [6,7,8] . It can describe the condition correlation between different knowledge using probability theory with graphics mode.…”