| Several frequency fluctuation accidents around the world caused by large scale power imbalance make consideration of frequency regulation issues urgent in steady-state operation analysis of power systems with large scale wind power, under the complex interaction circumstance of power sources, power systems and loads. This thesis proposed an improved fast decoupled power flow model considering the static characteristics of power systems, and optimized the static power frequency characteristics coefficients of generators. The proposed power flow model was applied in probabilistic power flow, and assessed the security risks of power systems. The main research work and achieved results are as follows.An improved fast decoupled power flow model considering the static characteristics of power systems was studied in this thesis. The static characteristics of generator and load were considered in the power flow model, which is more fitting the real operating features of power systems. When a disturbance occures in the power system, generator outputs and active load can varies with the change of frequency, which can reflect the primary frequency regulation process automatically. The model can reflect the secondary frequency regulation of the main regulation generators by parallel moving the static characteristics, when a greater disturbance occures in the power system. Block solving technology based on fast decoupled algorithm was adopted to improve the computational efficiency.The intelligent optimization of static power frequency characteristics coefficients of generators was studied in this thesis. Through a comprehensive and coordinating optimization of static power frequency characteristics coefficients of multiple generators under various operating modes, to minimize the absolute value of the system frequency offset. BP neural network based on Adaboost pretreatment is used to save computing time of repeatedly power flow calculation in every iteration in genetic algorithm optimization. Test results shows that the intelligent optimization method can get a satisfactory comprehensive and adaptive optimization result under various operating modes.A fast probabilistic power flow model considering static characteristics of power systems is studied, based on the proposed improved fast decoupled power flow model considering the static characteristics of power systems. Besides node voltages and branch flows, the scheme can obtain the probabilistic density function and cumulative distribution function of system frequency, which provides a usually ignored but of great importance desired variable as a study object of power system analysis, to discribe the operating features and distribution features of power systems more completely. Gram-Charlier series expansion based on culmulants was adpoted to calculate the probabilistc power flow. Test results show that, the proposed scheme has adequate accuracy, and obvious advantages in computing efficient compared to conventional probabilistic power flow based on Newton’s method. A variaty of disturbances were simulated to provide the probability distribution feature of node voltages, frequency and branch flows. The results demonstrate the scheme is fast and effective.The fast probabilistic power flow was applied to power system security risk assessment, to evaluate the operating conditions of node voltages, frequency and branch flows quantitatively by security risk assessment indices. An improved Von Mises method was adopted to avoid the shortages of Gram-Charlier series expansion method, and the results were satisfied. Special attention was paid to the power system security risk assessment when large scale wind power injected. Wind power multi-scenario was built and combined with complete probability method, which had solved the linearization model inadequacy problem when wind power fluctuates in wide range. Numerical examples have analyzed the power system security risk before and after wind power injected, and the effect to system frequency when different capacity of wind power injected. The results demonstrate the validity of the scheme. |
PDF Full Text Request