Power System Distributed Dynamic State Estimator

Phasor measurement unit can measure the phasors in the electromechanical transient process, which provide a new opptunity of power system dynamic monitoring and control. However, the PMU measurements experience random errors and bad data unavoidablely due to the sensor errors and disturbances. If the PMU measurements are applied directly without any treatment, the power system dynamic process may not be monitored accurately, even the wrong control scheme is likely be conducted by control system. In this dissertation, a power system dynamic state estimator in the electromechanical transient process based on PMU is studied. The main work of this dissertation is summarized as follows:(1) A framework of distributed dynamic state estimtimator is proposed. The dynamic state estimator for synchronous machines and the state estimator based on zero impedance characteristic are conducted in power plants and substations respectively. The estimation results are sent to the dispatch center where the power system state estimator is conducted. A novel approach for dynamic state estimator for synchronous machines in the electromachenical transient process based on PMU is proposed. The regulation effect of speed governors is considered. The model of dynamic state estimator for synchronous machines in the electromechanical transient process is built. The detailed calculation method of the system variance is given. The solving algorithm of the dynamic state estimator for synchronous machines based on the unscented Kalman filter using proportion symmetric sampling scheme is proposed. The simulation result shows that the precision of the proposed state estimator is higher than the method that the mechanical torque is assumed to be constant.(2) The PMU measurements experience bad data unavoidedly in the electromachenical transient process. Aiming at this problem, the robust dynamic state estimator for synchronous machines is proposed. A time-varying multidimensional scalar is introduced into the cubature Kalman filter, which regulate the measurements error in real time according to the innovation. Then, the regulated error is used to calculate the Kalman filtering scalar, which can accurately rectify the prediction of states under the condition that the PMU measurements experience bad data. The accurate estimation value of generator states can be obtained. If the time-vary multidimensional scalar is a non-diagonal matrix, the Kalman filter scalar will be a singular matrix, whose inverse matrix cannot be obtained. Aiming at this problem, the solving method is proposed. The detailed process of dynamic state estimator based on the robust cubature Kalman filter is presented. The simulation result shows that the robust dynamic state estimor can obtaine the accurate estimation results when PMU experiences bad data. (3) A novel approach for the substation state estimation based on PMU in the electromechanical transient process is proposed. In the approach, the redundancy of measurements is increased by taking the zero-impedance characteristic of circuit breakers as virtual measurements. At the same time, the state estimation model is built under the condition that the statues of the circuit breakers cannot be obtained in real time after a fault occurred in the system. By solving the model, the real status of circuit breakers can be obtained. The bad data identification method based on the non-quadratic rule state estimator is presented, and the method of choosing the threshold value is modified, which can identify the bad data effectively.(4) A state estimation method for the power system in the electromechanical transient process is proposed. Based on the direct solving method of machine-network interfacing, the calculation method of the pseudo measurement error of network voltage phasor transformed from the dynamic state estimation results for synchronous machines is given. The power system state estimator in the electromechanical transient process is proposed, in which the constraint of dynamic state estimator for synchronous machines is used to increase the accuracy.