基于复合本征噪声辅助分解的新型电力系统宽频振荡源定位OA

[Objective]With the gradual evolution of power systems toward a trend of"high share of renewables and high penetration of power electronic equipment",system oscillations have become more frequent and exhibit wide-frequency characteristics,posing greater challenges to operational security.[Methods]First,a compound intrinsic noise-assisted decomposition(CIND)method is introduced to perform modal decomposition on the wide-frequency information of each generator.Then,a targeted selection method for key wideband oscillation modes is developed based on correlation coefficients and energy.Subsequently,the dissipation energy flow corresponding to each generator is calculated from the key components to locate the oscillation source.[Results]The proposed method is validated on the WECC 179-bus test system and the NE 39-bus system with integrated wind farms.The results show that the localization accuracy remains close to 100%under noise levels of 100,65,55,and 45 dB,with single-machine computation time reaching the millisecond level.[Conclusions]Experimental results demonstrate that the method can accurately and effectively localize wideband oscillation sources under various noise conditions,showing good robustness and adaptability.