Studies On Forced Power Oscillation Caused By Strong Modal Resonance

The inter-area low frequency oscillation strongly affects the secure and stable operation of large-scale interconnected power grid. In recent years, several weak/negative damping low frequency oscillations and forced power oscillations have happened in China due to insufficient system damping or external successive and cyclic disturbances. Additionally, there are forced power oscillations caused by strong modal resonance. As one kind of first order modal resonance, strong modal resonance refers to the phenomenon that two close oscillation modes may move close together and interact as the system operating parameters or controller parameters experience minor changes. To fully understand and uncover the mechanism of forced power oscillation excited by one mode acting on another as a cyclical disturbance in the condition of strong modal resonance, the thesis titled by "forced power oscillation caused by strong modal resonance" carries out some fundamental research on oscillation mode pairing, identification of modal resonant points and resonant mode pairs, as well as forced power oscillation response analysis and type discrimination. The work has an important theoretical and practical significance in understanding low frequency oscillation mechanism, enriching and extending the analysis and control theory of low frequency oscillation, and improving the safe and stable operation of power system.Firstly, a mode pairing method based on perturbation theory is proposed, applying to the near and exact strong modal resonance condition. For a system after parameter perturbation, distinct or close modes can be approximated based on the accurate ones before the perturbation and their associated eigenvectors by using the perturbation theory. Through taking the modes approximations as an inter-medium, the accurate modes for the system before and after the perturbation can be paired. The method has a solid theoretical foundation and contains only one threshold that needs to be specified, so it’s easy for practical application. Simulation results show that the modes approximations have high precision and this method can used to pair the distinct and close oscillation modes even if the step of the parameter perturbation is large or the mode shape changes greatly after the parameter perturbation.Then, a method for identification and search of the potential resonant points and corresponding resonant mode pairs is proposed. Based on the obtained sets of resonant mode pairs and the corresponding perturbation parameters, a non-linear optimization model for identification resonant points and corresponding resonant mode pairs is built including small signal stability constraints. By utilizing the accurate and reliable mode pairing method, the model can be optimized and the existing resonant points and corresponding resonant mode pairs can be identified. Simulation results of the New England test system demonstrate that the resonant electromechanical-electromechanical mode pairs and electromechanical oscillation-control mode pairs can be effectively identified by the proposed method. Based on the identified strong modal resonant points and mode pairs, forced power oscillations excited by one mode acting on another as a cyclical disturbance can be observed botin in the single-machine infinite bus system and in the New England test system, which further proves that strong modal resonance can lead to forced power oscillations.Lastly,the response of forced power oscillation in multi-machine power systems is mathematically analyzed. The responses of resonance and beat frequency oscillation are obtained by further simplification based on their preconditions. Then, taking the differences in response components and their oscillatory characteristics as criteria, a method for discrimination of free and forced oscillation was proposed. The method has a merit of clear principle and complete criterion. Simulation results of the New England test system and a practical oscillation event demonstrate the validity of the response analysis of the forced power oscillation and the proposed method for oscillation type discrimination.In a summary, the thesis have gained some good results on oscillation mode pairing, identification of modal resonance points and resonance mode pairs, as well as forced power oscillation response analysis and type discrimination, which have built a good foundation for intensively studying the mechanism and the corresponding control strategies for the forced power oscillations caused by strong modal resonance. In the future, the demonstration of force power oscillation caused by resonant electromechanical-electromechanical mode pairs and the corresponding effective control strategy to avoid this kind of low frequency oscillation need to be analyzed and studied.