Distributed Model Predictive Control Of Wide-Area Power Systems

Along with the scale expansion of power systems, the power grids are gradually interconnected among cross-regions. The structure and dynamic behaviors of the interconnected power system becomes more and more complicated, which calls the importance of security and stability of power grids. Traditional SCADA systems focus only on the monitoring of system stable operation, but cannot do effective monitoring of the system dynamic behavior. The emergence and development of the Wide-Area Measurement System provides a new opportunity to the operation monitor, stability analysis and control of power system. Wide-Area Measurement System is built based on phase measurement and modern communication technology, and it serves the real-time control and operation of power system, and has broad prospects in power system application.However, although the grid interconnection and the application of Wide-Area Measurement System bring so many advantages, also bring lots of new challenges and difficulties, such as huge amounts of sampled data, frequent data transmission and long-distance transmission delays, etc. Though there have been some existing results on those issues, lots of problems still remain to be solved. Especially, how to design stabilization robust distributed controllers for Wide-Area Power System with quantization error and how to design the distributed controllers for the system with delayed states. In this thesis, the data quantization and states delay problems are considered, the distributed model predictive controller design of Wide-Area Power System is studied by employing Lyapunov stability theory and the linear matrix inequality technology. The main work this thesis is as follows:1. Modeling approaches of synchronous generator for power system are studied, and three practical models are produced. Considering the larger scale, complex dynamic and wide distribution of Wide-Area Power System, a distributed model is presented for Wide-Area Power System combining system structure.2. Due to the large amount data of wide area system and frequent transmission of huge amount data may cause congestion, a logic quantizer is introduced to quantify the data to reduce the length of packet data. Then the system model with quantization error is obtained, and a distributed robust model predictive control algorithm is presented.3. A distributed model for Wide-Area Power System with state delay is modeled based on synchronous generator model, and a distributed stable controller design method is proposed. By using Lyapunov stability theory and the linear matrix inequality technology an iteration distributed model predictive control algorithm is proposed, and the distributed controllers are obtained by solving a set of linear matrix inequalities iteratively. Finally, the proposed distributed algorithm is simulated on the two area four machine power system, the results verify the feasibility and effectiveness of the algorithm.4. A summarization is given, and the problems to be solved in the future are prospected.