CAREER: Managing uncertainties in renewable powered grids

这个NSF CAREER项目旨在开发算法来管理可再生能源发电固有的不确定性，以提供可靠和最低成本的电力系统运行。该项目将通过将可再生能源的不确定性充分纳入不稳定风险的在线评估和资源调度决策，为公用事业控制室带来革命性的变化。这将通过开发一种新的计算框架来实现，该框架将基于物理的电网建模、实时传感器测量、数值天气预报的伪测量与先进的机器学习和数据分析相结合，以实现管理电网运行中可再生不确定性所需的更高计算效率和准确性。该项目的智力优势包括提高可再生能源发电态势感知的新方法，评估瞬时不稳定风险，协调资源调度以减轻可再生能源不确定性的影响。该项目的更广泛影响包括对可再生能源电网不确定性管理基本理论的创新，以及对劳动力管道的改进，使其能够顺利过渡到100%脱碳电力系统。目前的先进技术，即缺乏海上风力发电的态势感知，通过离线研究确定在线应用的传输稳定裕度，以及静态/动态（容量）储备，不能充分捕捉具有高不确定性的可再生能源电网的新特征。该项目将通过以下方式弥补这些差距：1)通过提供融合基于物理的天气预报和深度学习方法的新框架，提高系统操作员的态势感知能力，以改进海上风力发电预报；2)通过开发一种近似系统动力学的计算数据分析算法，实现实时暂态不稳定风险评估；3)利用动态能源储备技术对运行储备进行重新设计，以纳入时空相关的可再生能源不确定性。这些基本理论和技术将公用事业控制室的风险管理科学从具有确定性实践的离线研究转变为数据驱动的风险意识在线解决方案，从而导致系统运营商做出明智的决策并及时采取缓解措施，以防止连锁事件的发生。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This NSF CAREER project aims to develop algorithms to manage the inherent uncertainties of renewable generation in order to provide reliable and least cost operations of electric power systems. The project will bring transformative change to utility control rooms by fully incorporating renewable uncertainties into online assessment of instability risks and decision-making of resource dispatch. This will be achieved by developing a novel computational framework that combines physics-based power grid modeling, real-time sensor measurement, and pseudo measurement from numerical weather predictions with advanced machine learning and data analytics to achieve higher computation efficiency and accuracy required to manage renewable uncertainties in grid operations. The intellectual merits of the project include novel methodologies to enhance situational awareness of renewable generation, assess transient instability risks, and coordinate resource dispatch to mitigate the impact of renewable uncertainties. The broader impacts of the project include innovations to fundamental theories of uncertainty management in renewable energy powered grids and improvements to the workforce pipeline enabling the smooth transition to 100% decarbonized electricity systems. Current state of the art technologies, i.e., lack of situational awareness of offshore wind generation, determining transmission stability margin via off-line studies for online applications, and static/dynamic (capacity) reserve, do not adequately capture the new features of a renewable powered grid with high uncertainties. This project will bridge these gaps by 1) enhancing system operators' situational awareness through providing a new framework to fuse physics-based weather prediction and deep-learning methodologies for improved offshore wind generation forecasting; 2) enabling real-time transient instability risk assessment by developing a computational data analytic algorithm to approximate system dynamics; and 3) redesigning the operating reserve via the dynamic energy reserve technology to incorporate spatially and temporally correlated renewable uncertainties. These fundamental theories and technologies move the science of risk management in the utility control room from offline study with deterministic practices to data-driven, risk-aware online solutions, leading to informed decision-making and prompt mitigation actions by system operators to prevent cascading events.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.