To eliminate the effects of parameter dimensions in the absolute eigenvalue sensitivity and account for the effects of probability distribution of wind speeds
a cumulative relative eigenvalue sensitivity (CRES) index and associated computing method is proposed. Based on CRES
impacts of various ancillary control strategies on damping torsional oscillations in the doubly-fed induction generator (DFIG)-based wind turbines can be quantitatively assessed. CRES is the sum of weighted relative eigenvalue sensitivity per se. By introducing relative eigenvalue sensitivity
the dimension of parameters can be removed. On this basis
by dividing wind speeds into several zones according to rotor operating states
the average wind speed of each zone is selected as the typical wind speed. And thus the weight of relative eigenvalue sensitivity for each zone can be determined. The method is verified through time-domain simulations. It is demonstrated that the CRES index can effectively identify key parameters
and re-tuning them can reduce the adverse influence of ancillary controllers on the torsional damping and decrease the torque impulse in the drive-train system. This work is supported by National High Technology Research and Development Program of China (863 Program) (No. 2014AA051901) and State Grid Corporation of China (No. KTB11201301545).