基于类噪声信号的等值负荷惯量两阶段辨识方法

负荷惯量的准确、在线估计是电力系统电压和频率稳定分析的重要基础。为此，提出了一种基于类噪声的负荷惯量辨识方法，基于系统中时刻存在的类噪声信号跟踪等值负荷惯量的时变特性。考虑到类噪声条件下惯量参数可辨识性较差，采用两阶段辨识的思路，首先对负荷的电磁参数进行辨识，然后基于第一阶段辨识得到的状态变量估计出惯量、转矩系数等机电参数。仿真算例和实测数据分析表明，所提负荷惯量辨识方法能够适应不同变转矩情形下负荷惯量的类噪声辨识需求；相较于恒转矩辨识方法，该方法能够更加准确地反映实际负荷的机电暂态特性，得到稳定、可靠的惯量辨识结果。

Ambient signal-based two-stage equivalent load inertia identification method

Accurate online estimation of load inertia is an important basis for power system voltage and frequency stability analysis. For this reason, an ambient signal-based load inertia identification method is proposed, which aims to track the time-varying characteristics of equivalent load inertia by ubiquitous ambient signals in system. Considering the poor identifiability of load inertia under ambient signals, a two-stage identification idea is adopted. To be specific, the load electromagnetic parameters are identified firstly, and the electromechanical mechanical parameters including inertia and torque coefficients are subsequently identified based on the state variables estimated in the first stage. The analysis of simulation cases and field measurements shows that the proposed method is able to meet the load inertia identification requirements under different variable torque conditions. Besides, compared with constant torque identification methods, the proposed method can accurately reflect the electromechanical characteristics of actual power loads, thus obtaining stable and reliable inertia identification results.