2025/07/01 by Lin, Pyei Phyo, Wächter, Matthias, Tabar, Mohammad Reza Rahimi +1
Computer Science · #Computer Science #Information and General Works::003: Systems Theory #Stochastic processes | Wind energy #Technology::620: Engineering #Technology::621: Applied Physics
paper · doi:10.15480/882.15314
The power conversion process of a wind turbine can be characterized by a stochastic differential equation (SDE) of the power output conditioned to certain fixed wind speeds. An analogous approach can also be applied to the mechanical loads on a wind turbine, such as generator torque. The constructed SDE consists of the deterministic and stochastic terms, the latter corresponding to the highly fluctuating behavior of the wind turbine (power output of a wind turbine or wind speed). Here we show how advanced stochastic analysis of the noise contribution can be used to show different operating modes of the conversion process of a wind turbine. The parameters of the SDE, known as Kramers-Moyal coefficients, are estimated directly from the measurement data. Clear evidence is found that both continuous diffusion noise and discontinuous jump noise are present. The difference in the noise contributions indicates different operational regions. In particular, we observe that the jump character or discontinuity in power production has a significant contribution in the regions where the control system switches strategies. We find that there is a high increase in jump amplitude near the transition to the rated region, and the switching strategies cannot result in a smooth transition. The proposed analysis provides new insights into the control strategies of the wind turbine.