Study On Power System Transient Stability Prediction Method Based On Responded Data

With the continuous development of large-area interconnected power grid, safe and stable operation of power grid is facing new challenges, so the power system transient stability analysis methods have higher requirements. Meanwhile, with the rise of wide area measurement system, it provides power system transient stability analysis method with a new direction. Based on wide area measurement system,this paper studies the methods of online-identification and decision-making for power system transient stability analysis, mainly as follows:(1) It studies the dynamic characteristics of transient stability process for both simple system and multi-machine system, and conducts in-depth analysis of theirs instability mechanism and stability criterion.(2) It proposes an improved least square support regression algorithm to predict the trajectory of the power angle based on the wide area measurement system. Compared with the original algorithms, the algorithm makes significant improvement on both prediction time and prediction accuracy.(3) Based on the change law of power angle trajectory dispersion degree after the disturbance, it raises the difference of standard deviation between sub-sequences as transient stability criterion with the use of phase space reconstruction. And it also forms a new power system transient stability prediction method based on responded data, which can judge the instability condition of power system much earlier.The paper carries out simulation with IEEE standard examples, and the results demonstrate the effectiveness and accuracy of the proposed method. The method can be used to make quick and accurate identification for power system transient stability with unknown parameters and inaccurate models.