Power System Transient Stability Prediction Based On Wide-Area Measurement System

With the realization of power grid interconnection, smart grid construction, power generation capacity increasing, large capacity, long distance transmission system operation, the extensive usage of power electronic devices and power market mechanism introduction, the stability of power system in China is becoming prominent increasingly, especially the transient stability of the system. Therefore, prediction and judgment on power system transient stability accurate and fast is imperative. The birth of wide area measurement system based on synchronized measurement technology provides strong guarantee for accuracy and speed of transient stability prediction algorithm.In this paper, the extended equal area (EEAC) algorithm is used as the theoretical basis to predict the transient stability of power systems. In the process of dividing cluster, this thesis computes the spatial distance between the generators using the norm theory, proposes a clustering index; in the equivalent process, through the analysis of the governor and prime mover model, this paper gets forecast model of mechanical power, breaks a prerequisite that is, assumption mechanical power is constant. Under the premise of Prediction speed is not affected, effectively overcomes the drawback of conservative when uses direct analysis method. Finally, through acquisition the data from wide area measurement system, using the least squares method calculate various parameters of the prediction model.This paper also studies the configuration of synchronized phasor measurement unit (PMU),defines the fewest number of PMU and system measurement redundancy as large as possible as the goal, then seeks the restriction condition which guarantees observability under N-1contingency set when system perturbation, finally get the method of optimal PMU placement based on transient considerable. This paper application of Power System Analysis Software Package (PSASP) calculates the typical example, obtains the needed transient stability prediction data. Algorithms, which proposed in this paper, are realized using MATLAB, and simulated on a typical example, the simulation results demonstrate the accuracy as well as the validity of the proposed method.