Study Of Parameter Identification Of Generator Units And Under Frequency Load Shedding

Parameter identification of generator units and under frequency load shedding are two vital issues in grid analysis. How to improve the accuracy of parameter identification and set appropriate parameters to achieve load shedding as little as possible with quick recovery of the frequency has been a popular research subject in academic and engineering circles. Directed against these two problems, an exploration and study has been done in this thesis. The contents are described as follows:1.In this thesis, a new approach for generator and excitation system parameter identification based on PMU grid connected operation data was presented. Firstly, the characteristics of grid connected operation instant-side terminal voltage and current have been analyzed, which to determine the validity on the unit identification. Then the generator and excitation system identification model was determined by the characteristics of PMU data. Finally, parameter identification of the generator and excitation system was achieved by the appropriate treatment for each model combined with particle swarm optimization algorithm. Meanwhile, considering the saturation problem of a generator, a new saturation correction strategy was proposed, which by introducing a saturation correction vector to quantify the realization of saturation effect. Numerical examples illustrate the practicality and efficiency of the strategy referred to the saturation correction2.In this thesis, a new approach for governor system parameter identification based on the measured frequency curve and PSASP simulated frequency curve was presented. Firstly, an input-output model of governor system was driven; then the simulated frequency curve of the disturbed power system was acquired by directly invoking PSASP electromechanical transient simulation module through an interface design. Governor parameters of generation system were iteratively adjusted to optimally match the simulated frequency curve with the measured frequency curve using PSO algorithm, thus governor system parameter identification based on the measured frequency curve had been realized. Simulation results of an actual power system illustrate the feasibility and efficiency of the proposed method.3.In this thesis, a new frequency security and stability evaluation standard was presented, which a detailed analysis and elaboration was made in the case of the low-frequency and high-frequency. Meanwhile, a novel optimization model for UFLS considering the importance of load and the sensitivity of frequency recovery and volatility on regional load shedding is built in the system for the low-frequency situations. Based on the classification of power load, the UFLS scheme corresponding to each type of load is set through the model established. With ensuring the global co-ordination for load shedding, while the minimum amount of both system load shedding and frequency offset are optimization objectives, proportion of each round of load shedding for each UFLS scheme is optimized based on the combination of electromechanical transient simulation and particle swarm optimization algorithm, eventually achieving the optimal under frequency load shedding scheme. Simulation results of an actual power system illustrate the feasibility and efficiency of the proposed method.