New Distributed Power System Dynamic State Estimation Algorithm Based On PMU

The interconnection of power systems is an inevitable trend with the development of future power system, we require the EMS to control the system better. Dynamic State Estimation(DSE) is the pivotal part of EMS, the algorithm of DSE is related to the speed and precision in calculating directly. With the aid of an efficient DSE, a system operator can be allowed more time in making control decisions such as economic dispatch, security assessment, or other related functions. For a large-scale power system, the conventional centralized state estimation methods have reached a stage in which substantial improvements in either speed or numerical robustness are unlikely. So we must open up new DSE methods to adapt to the development of future power systems.Phasor Measurement Unit(PMU) brings a new chance to the scientific research. Now how to apply the technique to the electrical power system is the focus of the research. As the same time, to adapt to the development of future power system, the decentralized control techniques will be the possibility applied to it.In the thesis, we proposed the distributed dynamic state estimation methodbased on PMU, established the corresponding math model, after studying the conventional and hierarchical methods. The main idea of the algorithm is "multi-processor to multi-area" and "exerting the PMU" in order to improve the precision and computing rate. Following aspects and works are done in this paper:1. The conventional DSE algorithm based on Kalman theory isn't fit for employment in application through the deduction of its formulas. Analyzed several improved algorithms to show its disadvantage.2. The phasor measurement technology based on GPS is introduced briefly. And how to add these phasor measurements into DSE is analyzed from several aspects. The differences between the conventional SCADA and algorithm based on PMU with several measurements precision are given.3. Analyzed the distribute algorithm. It can quick the calculating speed through decomposing a large system into tow or more subsystems. These subsystems can be solved and made observability analysis solely.4. Combined with PMU and parallel technology, the new distributed DSE based on PMU algorithm was given.By using the data of PMUs, It really achieved parallel process, avoided the relay by exchanging data of the previous one. Combined with the methods of processing the data before filtering step, weighing the elements of Jacobian matrix and so on, the new algorithm can calculate fast and accurately.After calculating the IEEE-5 and IEEE-14 system by the software of Visual Basic and MATLAB in this paper, The results indicated its advantage and efficient.