This paper proposes a comprehensive method to estimate non-synchronous inertia in an AC microgrid system by analyzing measured frequency transients. Non-synchronous inertia includes inertial support from various inverter-interfaced renewable energy sources and energy storage devices in the form of virtual or emulated inertia. In the proposed method, a multi-step approach is used to identify a matrix of parameters that estimates the overall non-synchronous inertial response. Firstly, motivations are drawn from the fundamental system requirements of adequacy in RoCoF and maximum deviation in frequency, for any credible contingency. A systematic classification of various contributors to the non-synchronous inertial response is also carried out. System-identification studies are then performed to identify various component-models that contribute to the non-synchronous inertia. The component models include a faster RoCoF-based response model, a relatively slower frequency deviation based model, and in some cases, a slow, integral-of-frequency based model. Lastly, the concept of releasable kinetic energy and non-linear estimation is used to identify the capability of the virtual inertia sources and improve the accuracy of the overall estimate. The developed methodology is tested in a microgrid network hosting a mix of renewable energy generations and battery energy storage systems. The estimate of non-synchronous inertia will help system operators to accurately estimate the maximum frequency deviation and the time to reach the maximum deviation for any credible contingency. The estimate can also help in deciding the type of virtual inertia that will best maintain frequency stability in the microgrid system.

中文翻译：

---

交流微电网中非同步惯性的估计

本文提出了一种通过分析测量的频率瞬变来估计交流微电网系统中非同步惯量的综合方法。非同步惯性包括来自各种逆变器接口的可再生能源和储能设备以虚拟或模拟惯性的形式提供的惯性支持。在所提出的方法中，使用多步方法来识别估计整体非同步惯性响应的参数矩阵。首先，动机来自于 RoCoF 的充分性和频率的最大偏差的基本系统要求，以应对任何可信的意外情况。还对非同步惯性响应的各种贡献者进行了系统分类。然后执行系统识别研究以识别导致非同步惯性的各种组件模型。组件模型包括更快的基于 RoCoF 的响应模型、相对较慢的基于频率偏差的模型，在某些情况下，还包括基于频率积分的缓慢模型。最后，利用可释放动能和非线性估计的概念来识别虚拟惯性源的能力，提高整体估计的准确性。开发的方法在托管可再生能源发电和电池储能系统的微电网网络中进行测试。非同步惯量的估计将帮助系统操作员准确估计最大频率偏差和达到任何可信突发事件的最大偏差的时间。