Dynamic task decomposition and allocation in CGF

Computer Generated Forces (CGF) is virtual entities of combat force, which can be generated and controlled by computers. Human Behavior Representation (HBR) in CGF is critical for combat simulation. Yet, despite the considerable progress in related research, the challenge of dynamic task decomposition and allocation has remained largely unaddressed. This paper describes the task decomposition strategy and principles, presents related algorithms based on the task teams. Through taking the agent´s experiences, abilities, and individual states into account, paper designs and implements the task allocation algorithm based on utility theory. The experimental results showed that our algorithms were effective, applicable and can reflect the task decomposition and task allocation process in domain of combat simulation.