基于数据挖掘技术的电力网络状态估计研究-控制理论与控制工程专业论文.docxVIP

沈阳理工大学硕士学位论文 沈阳理工大学硕士学位论文 Abstract Power system state estimation aims to improve data accuracy using redundancy of measurement system, which automatically eliminates error messages caused by random disturbance, and to estimate and forecast real operation of system. The accuracy of the estimation is the basic requirements of power system state estimation. Bad data detection and identification is one of the important functions of power system state estimation, which plays an important role in the accurate result of state estimation. For transmission network, because of the large number of measurement devices and measured data and redundancy, Traditional detection and identification was mainly focused on transmission system, but residual pollution and residual submerged existed. Therefore, how to detect and identify bad data accurately in state estimation is an inevitable research topic. For distribution network, because of the wide distribution and large number of nodes, and lack of measurement devices and measured data, which needs to generate pseudo measurement data to add load data to estimate state. Owing to the accuracy of pseudo measurement generated by transformer capacity is quite different from measured data, which results in inaccuracy or even no convergence of DSE. Therefore, we have to find a new method to supplement to the pseudo measurement which is close to measured data. For the problems mentioned above, in order to improve the accuracy of the results of state estimation, this paper takes a node load data for research object and uses K- means algorithm which combines validity index to extract daily load characteristic curve, which, then, is to be used for bad data detection and identification in transmission system and pseudo measurement complement in DSE. Eventually, we programm with MATLAB, then make the simulation analysis, the result shows that this method is effective. Key words：State estimation，Data mining，Bad data detection and identification， K