Subsynchronous oscillation random time-varying mode identification method

Abstract

The invention discloses a subsynchronous oscillation random time-varying mode identification method. Aiming at the problem that an error of mode identification result is relatively large due to an end effect, an improved Hilbert-Huang transformation based on a mirror image extension method is proposed to identify a random time-varying subsynchronous oscillation mode, first, symmetry points are determined at left and right ends of a signal, the signal is then extended, fitting of envelope lines adopts extreme points after extension as interpolation points to perform cubic spline interpolation,and then the local mean of the upper and lower envelope lines is calculated, so that the fitting of the upper and lower envelope lines is more accurate, the end effect is notably improved, and the feasibility and effectiveness of the method are verified by the identification of stable signals, non-stable signals and measured signals.

Description

Technical field [0001] The invention belongs to the technical field of electrical engineering, and specifically relates to a subsynchronous oscillation random time-varying mode identification method. Background technique [0002] As an important aspect of power system stability, subsynchronous oscillations have received extensive attention and research since the 1970s. With the continuous development of large-scale wind power and large-scale photovoltaic power generation bases, the grid-connected capacity continues to increase, and large-capacity power electronic technologies are widely used, and subsynchronous oscillations have shown new characteristics. Affected by natural factors such as wind speed, the production of new energy power sources has relatively large random volatility. This random volatility causes random drift in the operating point of the power system. The range of operating conditions is larger and the changes are more complex. The influence on the subsynchrono...

AI Technical Summary

Problems solved by technology

On July 1, 2015, more complex large-scale double-fed and direct-drive wind turbines sent out through weak AC and Tianzhong UHVDC systems occurred in the Hami area of ​​Xinjiang in my country. According to on-site monitoring data, wind power The oscillation frequency of the generated sub-synchronous current changes with time and propagates in the AC system of different voltage levels, which finally induces the shafting torsional vibration of the thermal power unit, which in turn causes the unit torsional vibration protection to start, and the three thermal power units trip

In this way, the envelope at the endpoint deviates from the true value, the local mean curve is inaccurate, and the screened out Intrinsic Mode Function (IMF) is also inaccurate

The IMF error obtained by this decomposition will affect the subsequent EMD process, and the error at the endpoint will gradually pollute inward, distorting the decomposition result

[0006] To sum up, the subsynchronous oscillation of large-scale new energy power system presents the characteristics of random time-varying amplitude and frequency. Therefore, it becomes difficult to accurately identify the mode of subsynchronous oscillation, and the Hilbert-Huang transform (Hilbert-Huang transform) Huangtransform, HHT) can accurately extract information such as frequency and amplitude from non-stationary signals, but HHT has endpoint effects, which will cause large errors in the identification results at the boundary

Images