Multi-agent Planning Techniques In The Air Force In The Battle Of Planning Applications

As an approach to cooperative problem solving, multi-agent planning is aiming at completing a problem-solving task by providing a global coherent plan which can be executed concurrently. The research field mainly includes the representation of plan, the method of planning, the multi-agent planning architecture, the technologies of coordination and etc.The purpose of this thesis is to create a sequence of actions of the cooperative operations using the technologies of multi-agent planning. The thesis focuses on the application-related technologies in multi-agent planning research field: the model of representation for plan based on a plan ontology, reasoning about location, plan basic organization structure model and the algorithm of plan merging.After introducing the aim and the method of constructing a plan ontology in detail, the thesis presents a plan ontology, and based on which a plan generating algorithm is further proposed using the technology of reasoning about location. Then the thesis describes the model of plan basic organization structure that is suitable for air campaign planning, Under this frame, the algorithm of plan merging of multi-agent planning is implemented and its chief process is expressed in C++-like language. Finally, a simple example is adapted to explain how to make use of these technologies to create a sequence of actions in cooperative operations in air campaign planning.The main research achievements and innovations of this thesis are the proposing of a more complete and better formalized plan ontology, and a more suitable plan merging algorithm for air campaign planning based on the representation model of such plan ontology.