Research On The Emergency Control Of Hybrid Power System Based On Finite Automata Network

The high speed development of modern power systems makes the scale of the power network larger and larger, i.e., its architecture tends to be increasingly complex and its stability and safety are becoming a formidable challenge when various new technologies are introduced, Accordingly it turns essential to design more efficient controllers to coordinate and optimize the operation status of power system. And how to realize the optimizing schedule and control of various component of such very large scale hybrid electric power system remains an urgent issue. Firstly the safety problem of electrical power system is analyzed in this thesis. Both the background and the relevant problems concerning emergency control of frequency and/or voltage of the electrical power system are discussed. In particular its hybrid natures are brought to light. Secondly the general theory of hybrid control system, essential for modeling the emergency control system of the power system, is presented, which includes the 9-tuple architecture of hybrid control system, the approaches for synthesis of hybrid control system, and the practical applications as well. The hybrid nature of power system and the necessity for the use of hybrid control system methodology to model the emergency control system of the power system are emphasized. Next some fundamentals of finite automata, utilized for modeling hybrid control system, are given, e.g., the 5-tuple architecture, the state diagram representation and its fundamental properties, together with its typical engineering applications. Learning automata, a special type of finite automata, is also discussed afterwards. The celebrated reinforcement scheme for learning automata is analyzed in more detail, i.e., three well-known schemes are given and the operating process of learning automata is concretely analyzed with an illustrative example. In summary, an emergency control model for hybrid power system is proposed, which is the original contribution of the present investigation. In addition, the model is applied to both frequency and voltage emergency control problems. Combined with some test-bed problems, the model is carefully analyzed and verified and it is shown that the model suggested in this paper outperforms conventional ones.