Deterministic Learning‐Based Fast Detection of Sub‐Synchronous Oscillations in Wind Power Grid Connection

With the increasing integration of large‐scale renewable energy sources into the power grid, the issue of sub‐synchronous oscillation (SSO) has become more pronounced, posing a significant threat to the safe and stable operation of the power system. This paper proposes a new method for the early and rapid detection of SSO using deterministic learning theory. A comprehensive database of oscillation patterns is constructed by simulating a variety of conditions that can induce grid oscillations, including wind speed and control parameters. A learning estimator is then developed to identify stable and oscillation dynamics. Furthermore, a knowledge bank for SSOs detection and isolation is also established, using the minimum residual principle, oscillation patterns matching those in the knowledge bank can be detected. Then, based on an example of grid‐connected sub‐synchronous oscillations in wind turbines are utilized to show the effectiveness of the proposed method. The detection accuracy of SSO is calculated as 94%, and the detection time in this paper is reduced to about 0.023 s.