Study On Dynamic Harmonic State Estimation Technology Based On Kalman Filter

In recent years, with the development of power electronics technique and the use of non-linear loads, more and more harmonics are injected into the network. Harmonic becomes an important element of impact on electric quality in power system, so it is necessary to monitor power system harmonic. Harmonic State Estimation(HSE) provide system harmonics voltage level to power board according to the system topology, key harmonic voltage and current measurement, and then determine the system harmonic information.The modern measurement technology is prior condition of the HSE, meanwhile it brings some advancement for the relevant problems about measurement variables, estimation models, solution algorithm and observation analysis. Firstly, the paper introduces the basic theory of Beidou Navigation Satellite System and the harmonic measurement system, and it brings up measurement models and dynamic state estimation methods(DHSE) by selecting the synchronism measurement value of bus voltages, bus influx currents and branch currents as measurement variables and the bus voltages in the whole net as status variables. Secondly, A SVD-Kalman filter is proposed to solve the non-quasi state dynamic state estimation.The simulation and procedures are used to verify it. Finally, detection methods of the bad data and for DHSE are proposed. The major work and conclusions are as follows:A dynamic harmonic state estimation mode is formulated based on Kalman filter. The main target of the Kalman filter is the optimal estimation of the state variables from measurements corrupted with noise, by minimizing the square of the expected error between the values of the actual and estimated system states. There are two kinds of adaptive Kalman filters. The first one is adaptive method to determine the noise covariance matrix Q and measurement noise variance R . The second one is adaptive fading Kalman filter. As the system transition matrix and measurement matrix have the feather of multidimensionality and the computer roundoff may degrade the performance of the Kalman filter, a fading kalman filter with offline Q and R determind is advanced.The IEEE-14 bus test system simulation and procedures are used to verify it.While the power system is under quasi-static state and no bad data exists, comparison shows the precision is approximate for DHSE and SSHE (static harmonic state estimation), but while the system is not under the quasi-static state, the Kalman method precision is not good anymore; while there are bad data measurements, the Kalman filter shows it has the ability to detect bad data. A hybrid estimation algorithm called SVD-Kalman is proposed to solve the system under transient state problem.While the harmonic state estimation computation procedure begins, SVD procedure is enable. While the system is under quasi-static state, a fading Kalman filter is used to fulfill the estimation.At last, the bad data detection theory and its influence to the dynamic state estimation are introduced, rN detection method and the suddenly-change detection method are used to SVD-Kalman algorithm, and the basic criterion for automatic selection of detection is included.