Study On WAMS Based Power Systems State Estimation And Stability Assessment

Wide Area Measurement System (WAMS) realize the measurements for the complete dynamic process of power systems in three different time frames which includes steady snapshot, fault (perturbation) scenarios and electromechanical swing. The WAMS measurements implicate power system information from steady state to dynamic state. It is still a hard task that how to extract the useful information from the raw WAMS data. The dissertation focuses on data mining and information extraction for power system steady state, fault scenario and transient stability, after eliminating the noise and bad data of raw WAMS data. They are mainly including:Firstly, a novel hybrid nonlinear state estimator including PMU node voltage phasor and branch current phasor measurements is proposed. The proposed estimator adopts the PMU current phasor measurements directly (after necessary coordinates transform), and incorporates them with PMU voltage phasor measurements and conventional SCADA data to solve the nonlinear equations iteratively. The error covariance of indirect measurements caused by coordinates transformation is analyzed in detail, and is compared with those caused by SCADA measurements transformation. The simulation results demonstrate that the proposed estimator has higher estimation precision, better convergence capability and less iteration numbers. To further reduce the convergence time, the fast decoupled algorithm of the proposed method is realized as well.Aiming at the measurement errors of WAMS dynamic data, a distributed dynamic state estimator based on WAMS during power system electromechanical transient process is proposed. The proposed estimator chooses the rotor angle and electrical angular velocity as the generator variables to estimate. The generator output power measured by PMU is used to decouple the generator rotor movement equation and the outer network equation. And the linear Kalman filter based dynamic state estimator mathematical model is presented. The WAMS measurement noise and dynamic model noise are analyzed in detail. The total flow chart and the bad data detection and elimination approach are given as well. The simulation results indicate that the proposed dynamic estimator has high estimation precision and fast computational speed; therefore, which can satisfy the real-time requirements of the dynamic monitoring and control system.For the information extraction for fault scenarios, the measured positive sequence phasors of branch currents is adopted as characteristic variables, and a novel disturbance recognition algorithm based on WAMS is proposed. Utilizing the principle of superposition, the proposed method injects the relative currents into the presupposed event node to get the estimated disturbance components of the branch currents in the disturbance network. The PMU measured vector consists of PMU measured positive sequence phasors of branch currents, and the presupposed event vector consists of estimated disturbance components of the branch current phasors for each presupposed event. Then, the 2-norm of the distance vector between the measured vector and each presupposed event vector is calculated and the minimal 2-norm indicates the true event. The simulation results show that the proposed disturbance recognition algorithm can recognize and match the fault (disturbance) scenarios correctly and rapidly, which is useful for the remedial control system.Aiming at the transient stability information extract for estimated WAMS trajectories, the dissertation discuss some problems in the application of EEAC (extended equal area criterion). After simulation and analysis, it is pointed out that the frozen parameter method using admittance online identifying technology is not a proper way to calculate the stability margin in large systems with measurement noise, and the curve extrapolation technology should be a better alternative. A novel integrated rotor angle criterion is proposed to identify the true critical cluster when observed snapshot time is chosen properly. And an integrated criterion is proposed to judge whether the DSP is the true DSP which resulted in losing stability. Finally, the flowchart of transient stability information extraction based on WAMS is proposed, and the simulation results demonstrate that the proposed method is effective. It can judge transient instability even if the characteristic of instability is not obvious.