Collaborative Research: Advancing Robust Control and State Estimation of Converter-Based Power Systems

未来的电网是国家最关键的基础设施，由于可再生能源的大规模整合，未来的电网将极难管理。该项目中提出的战略得到了新的电网技术(风电场/太阳能发电场中基于变流器的资产和高频传感设备)的帮助，这些技术的开发和部署允许以较少的开销实施新的实时控制理论算法-同时保证电网的稳定性和弹性。这一领域的文献讨论了各种科学研究问题，但大多采用简化模型，不能充分反映未来电网的实时运行。该项目通过开发一套新的实时算法来解决这一科学空白，从而使以可再生能源渗透率高为特征的未来电网的运行更加稳健。这些控制算法可以在全国范围内由电网运营商实施。该项目还将包括：a)为圣安东尼奥一所低收入、少数族裔占多数、仅限女性就读的高中主办一次关于可再生能源系统的外联讲习班；b)组织一次技术行业讲习班，展示爱荷华州创造的算法；c)在德克萨斯大学圣安东尼奥分校和爱荷华州立大学的课程中传播创造的科学方法。特别是，控制理论文献解决了过多的电网挑战，重点放在线性化的微分方程式模型上，从而消除了代数约束(即潮流)。这与更现实、更复杂的非线性微分代数方程(NDAE)模型形成了鲜明对比。将工作点周围的网格模型线性化并消除代数约束已被证明是一种可靠的策略-在复杂性和可管理性之间进行权衡。然而，随着断断续续的可再生能源越来越多地将电网推向极限，由于对风能或太阳能的预测不佳，电网的物理状态有可能逃离运行区。对于以高度不确定的可再生能源为特征的电网来说，NDAE的控制取代了线性微分方程式模型是非常有益的。这保证了电网在较大运行条件下的稳定性。鉴于现有电力系统模型的局限性以及缺乏电网NDAE控制和动态状态估计的理论基础，本项目将：1)结合常规电机和各种基于变流器的技术，建立具有物理代表性的电力系统NDAE模型；2)研究动态状态估计的一般理论和考虑由高阶NDAE建模的电网的不确定性的鲁棒反馈控制算法；3)获得可在电网控制中心实现的易于计算的例程。所建立的理论基础在广域控制、基于变流器的控制、集中式和分散式控制以及鲁棒动态状态估计等方面都有应用。这项研究对保证现代和未来电力系统的可接受性能至关重要，并将推动电网控制研究的最先进水平。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Future power grids, the nation’s most critical infrastructure, will be extremely difficult to manage due to large-scale integration of renewable energy resources. The strategy proposed in this project is aided by new grid technologies (converter-based assets in wind/solar farms and high-frequency sensing devices) that are developed and deployed to allow new real-time control-theoretic algorithms to be implemented with little overhead---while guaranteeing grid stability and resilience. The literature in this area had addressed various scientific research questions, but mostly adopted simplified models that cannot adequately capture the real-time operation of future grids. This project addresses this science gap by developing a new set of real-time algorithms, leading to a more robust operation of future power grids characterized with high penetration of renewable energy resources. These control algorithms can be implemented by grid operators throughout the nation. The project will also include: a) hosting an outreach workshop on renewable energy systems for a low-income, minority-majority, and female-only high school in San Antonio; b) organizing a technical industry workshop that showcases the created algorithms in the state of Iowa; c) disseminating the created scientific methods within the curricula at the University of Texas at San Antonio and Iowa State University.This project aims at modernizing grid control methods which has traditionally relied on linear systems theory. In particular, the control-theoretic literature addressed a plethora of grid challenges with a focus on linearized, differential equation models whereby algebraic constraints (i.e., power flows) are eliminated. This is in contrast with the more realistic, complex nonlinear differential algebraic equation (NDAE) models. Linearizing grid models around operating points and eliminating algebraic constraints have proven to be a reliable strategy---a trade-off between complexity and tractability. Yet as grids are increasingly pushed to their limits via intermittent renewables, their physical states risk escaping operating regions due to a poor prediction of wind or solar. In lieu of linear differential equation models, control of NDAEs is highly beneficial for grids that are characterized by highly uncertain renewables. This guarantees grid stability for larger operating conditions. Given the limitations of present power system models and the lack of theoretical foundations for control and dynamic state estimation of grid NDAEs, this project will: 1) create a physically representative NDAE model of a power system with a mix of conventional machines and a variety of converter-based technologies; 2) investigate a general theory of dynamic state estimation and robust feedback control algorithms that consider the uncertain nature of power grids modeled via higher-order NDAEs; 3) obtain computationally tractable routines that can be implemented in control centers of power grids. The created theoretical foundations have applications in wide area control, converter-based control, centralized and decentralized and robust dynamic state estimation. This research is critical to guarantee acceptable performance of modern and future power systems and will lead to advancing the state-of-the-art of grid control studies.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.