Study On Multi-Agent System Modeling In Drilling And Well Control Simulation

Various problems may happen in oil drilling. Especially, if the well loses control, serious accidents may occur. In order to master the well-control skills and keep the well safe when abnormal situation occurred while drilling, it is required to develop the well-control simulation system for training the workers.As the collaboration, concurrency and flexibility of the drilling simulation process before were poor, this thesis studies and proposes a new method for modeling the well control simulation system based on multi-agent system theory. The main contributions are as follows:(1)In view of the existing problems of the present well control simulation system, a new well control simulation system structure is established by multi-agent technology and the agent classification and the structure of each agent are discussed.(2)An effective method of knowledge representation based on relation model is used to describe the attributes and rules of each agent in the simulation system.(3)Primitive parameters in the conventional Knowledge Query and Manipulation Language (KQML) are extended and the extended language is implemented by C++ programming. (4) Each agent in the system is implemented and constructed by the object-oriented method, and an agent-based normal drilling well control simulation system is developed. The simulation results show that the multi-agent based simulation mechanism has better performances than traditional drilling simulators, and it also lays the foundation for the further implementation of the whole system.