Uncertainty Evaluation and Mitigation in Electric Power Systems

本研究的目的是发展计算上有效的方法来评估参数不确定性对潮流解和电力系统动态性能的整体影响。该方法集成了两种有前途的工具，即概率搭配方法和基于轨迹灵敏度的方法。这些工具允许选择一组模拟，从中可以找到不确定参数与感兴趣的输出之间的映射。该映射可用于研究关键不确定性的影响，并调整参数以减轻这些影响。智力价值：该项目的智力价值在于研究了允许对电力系统的大型动态模型进行不确定性评估的方法。该项目将开发选择稀疏模拟集的技术，这些模拟集可以在可能的参数值范围内进行性能评估。对轨迹敏感性的研究将寻求建立对动态模型对不确定性传播影响的基本理解。将提供对网络结构所起作用的见解，并开发利用该结构的工具。更广泛的影响：公用事业使用基于模拟的系统行为评估来做出具有巨大经济和可靠性影响的决策。然而，许多模型的参数是相当不确定的。本项目开发的技术将允许对这种不确定性进行有效的评估。发生大骚乱的风险将相应降低。鉴于所开发的方法对一般大型系统的广泛适用性，该项目将刺激各种教育机会。

The objective of this research is to develop computationallyefficient methods for evaluating the global impact of parametricuncertainties on power flow solutions and the dynamic performance ofelectrical power systems. The approach is to integrate two promisingtools, the probabilistic collocation method and trajectorysensitivity-based methods. These tools permit the selection of asparse set of simulations from which a mapping between uncertainparameters and the output(s) of interest can be found. The mappingscan be used to study the effects of critical uncertainties, and totune parameters to mitigate those effects.Intellectual Merit:The intellectual merit of the project lies in the study ofmethodologies that permit uncertainty evaluation for large dynamicmodels of electric power systems. The project will developtechniques for selecting a sparse set of simulations that enableperformance evaluation over the domain of possible parameter values.Investigations of trajectory sensitivities will seek to establish afundamental understanding of the influence of dynamic models onuncertainty propagation. Insights into the role played by thenetwork structure will be provided, and tools that exploit thatstructure developed.Broader Impact:Utilities use simulation-based evaluation of system behavior to makedecisions that have tremendous economic and reliabilityimplications. The parameters of many models are, however, quiteuncertain. The techniques developed in this project will allowefficient assessment of such uncertainty. The risk of largedisturbances will be reduced accordingly. Given the wideapplicability of the developed methods to general large-scalesystems, the project will stimulate diverse educationalopportunities.