递归随机子空间辨识方法在电力系统区间振荡模式估计中的应用

Application of Recursive Stochastic Subspace Identification Method to Power System Inter-area Oscillation Mode Estimation

本文用随机子空间辨识方法在电网不同工况下对其区间振荡模式进行估计。同时,针对现有方法在线估计时奇异值分解步骤次数过多的问题,提出具有递归形式的随机子空间辨识方法以大大降低运算量。最后,将两种算法分别在四机系统和美国东北部NPCC48机系统进行区间模式估计试算。结果表明,随机子空间辨识方法能同时利用暂态数据和类噪声数据准确估计区间低频振荡模式,而递归随机子空间辨识方法既保证较好精度,又能提高原有算法的效率,为在线模式追踪提供了可能。

Stochastic subspace identification(SSI)method is applied to estimate the inter-area oscillation modes(IOMs)under various operating states of power system in this paper.Meanwhile,in light of the too many steps in singular value decomposition(SVD)during an on-line estimation when using the existing methods,an SSI method in a recursive form(RSSI)is also proposed to reduce the computation load significantly.At last,both methods are examined on a four-machine test system and the Northeastern Power Coordinating Council(NPCC)48-machine system.Results show that the SSI method can accurately estimate the low-frequency IOMs through the utilization of both the ringdown and noise-like data.In comparison,the RSSI method can not only ensure a higher accuracy,but also improve the efficiency of the original algorithm,thus making the on-line mode tracking possible.