Research On Power Network Parameter Estimation Method And Its Application

With the increasing scale, power network emphasizes its automation and safety more than before. So EMS has the opportunity to take a more important role in the automation control and the safely operation. Network parameter is the essential data in EMS database. And nearly all the advanced application softwares in EMS need network parameter in their computing process. Generally, the network parameters in EMS database are thought accurate. Actally, sometimes they are not as accurate as thought. Somethings can cause mistakes to them, and we call that parameter mistake. Accurate parameter promise the accurate results from the advanced application softwares. So we need power network parameter estimation.Power network parameter estimation is composed of two part, parameter mistake identification and parameter value calculation.Innovation graph theory provides a new way for parameter mistake identification. In innovation graph parameter mistake identification method, the general branch model is built to to identify the parameter mistake, which transforms the effection of parameter mistake to adjunction potential source. This method is not complicated, and has loose demand to measurement redundancy. But this method does not take measurement error and forecast error into account, which may lead the identification to a failure.In this dissertation, the author improve innovation graph parameter mistake identification method. Both active power innovation and reactive power innovation are used to identify parameter mistake, which can increase the reliability of identification in the condition of taking measurement error and forecast error into account. In order to decrease the effections of measurement error and forecast error, data in serial time sections are used into innovation graph parameter mistake identification method.After identifying the parameter mistake, we should estimate its parameter value. In this dissertation, GATS method is used to calculate the parameter value, which is composed of GA (Genetic Algorithm ) and TS (Tabu Search). With the help of GATS, parameter value calculation can avoid the numerical problems that may happen in augmentation state estimation.At last, according to the power network parameter estimation method issued in this dissertation, the author write a parameter estimation program, which is used in real power system. The result prove its availability.This dissertation is supported by NSFC and SGCC.