Hybrid agent model to design real-time distributed supervising and control systems

The real-time distributed supervising and control systems have distributed physical entities that cooperate together to accomplish their local goals, make possible decisions, execute actions, and negotiate through the communication protocol to reach global criterion. For designing its simulation software, the hybrid software agent model is developed on the basis of deliberative and reactive agent concepts with some special characteristics. The proposed designing method for real-time distributed supervising and control systems as MAS mainly involves building an ontology model, formally specifying protocols and determining the real-time and deliberative behavioural rules based on the ontology. The reactive memory is introduced to the model for guaranteeing to respond urgent events in real-time. The deliberative part enables the agent to do complex tasks based on agent's mental states and input event sequence. An example is given to show the validity of the designing method.