Intelligent Agents for Modeling Warfare Entities Actions Using an Improved Contract Net Protocol

In order to solve the problems about simulation of intelligent warfare entities actions on dynamic battlefield environment because of the complexity of warfare system, intelligent-agents-based modeling approach is applied in this paper. Based on analysis on caste diagram of warfare entities, the mapping from warfare entities to respective agents is set up, by which the simulation infrastructure is designed. According to the real situation of warfare entities actions, an improved Contract Net Protocol (CNP), which has functions of conditions constraints and confirmations for some key military tasks in contrast to the original CNP, is proposed to describe multi-agent interactions. Agent Unified Modeling Language (AUML) sequence diagram is applied to express the improved CNP. The demonstration system proves that our approaches are feasible and effective, and shows that our model can be used to understand the external, complicated and intelligent military operations activities and can realize the dynamic warfare actions simulation.