With the proliferation of renewable energy and electric vehicles (EVs), there have been increasing uncertainties in power systems. Identifying the influencing random variables will reduce the effort in uncertainty modeling and improve the controllability of power systems. In this paper, a density-based global sensitivity analysis (GSA) method is proposed to evaluate the influence of uncertainties on islanded microgrids (IMGs). Firstly, the maximum IMG loadability evaluation model is established to assess the distance from the current operation point to the critical operation point. Secondly, the Borgonovo method, which is a density-based GSA method, is used to evaluate the influence of input variables on IMG loadability. Thirdly, to improve GSA efficiency, a modified Kriging model is used to obtain a surrogate model of IMG loadability, and Borgonovo indices are calculated based on the surrogate model. Finally, the proposed method is tested on a 38-bus IMG system. Simulation results are compared with those considering other methods to validate the effectiveness of the proposed method. Energy storage systems are considered to diminish the influence of critical uncertainties on IMG operation.

中文翻译：

---

考虑分布式能源整合的基于密度的孤岛微电网负荷能力全局灵敏度分析

随着可再生能源和电动汽车（EV）的激增，电力系统的不确定性日益增加。识别影响随机变量将减少不确定性建模的工作量，并改善电力系统的可控性。本文提出了一种基于密度的全局灵敏度分析（GSA）方法，以评估不确定性对孤岛微电网（IMG）的影响。首先，建立最大IMG负荷评估模型，以评估从当前运行点到关键运行点的距离。其次，使用基于密度的GSA方法Borgonovo方法来评估输入变量对IMG负载能力的影响。第三，为了提高GSA效率，使用了改进的Kriging模型来获取IMG可加载性的替代模型，和Borgonovo指数是根据代理模型计算的。最后，该方法在38总线IMG系统上进行了测试。将仿真结果与考虑其他方法的结果进行比较，以验证所提出方法的有效性。储能系统被认为可以减少关键不确定因素对IMG运行的影响。