CAREER: Risk-Averse Decision Making via Chance-Constrained Programming for Power Systems

NSF CAREER项目旨在开发规避风险的决策模型和方法，以管理高度不确定的大规模可再生能源的整合，使社会受益于更清洁、更可靠和更具成本效益的电力系统。由于生产的间歇性和有限的可预测性，可再生能源的增加给电力系统带来了重大挑战。在经济可靠地平衡电力供需时，传统的运行策略可能在这种不确定性增加的情况下是不够的。CAREER项目将弥合这一紧迫的不确定性差距，有效地整合可再生能源，从而有助于确保系统的可靠性和成本效益，这将给电力行业带来革命性的变化。该项目的智力优势包括开发新的规避风险的决策模型，以适应电力系统的性质和决策者的风险偏好，并开发新的计算高效的解决方案方法，为所有利益相关者和最终用户社区提供实用、有用的解决方案。该项目更广泛的影响包括：(i)通过指导本科生进行该CAREER项目的实践研究，将亚利桑那大学的研究和教育结合起来，并将研究结果与熟悉的应用联系起来，以激励下一代学生从事工程职业；（ii）纳瓦霍族能源挑战的可持续解决方案；（iii）在美国经济中具有重要社会意义的行业（如交通和金融）存在高度不确定性的情况下，向其他应用领域转移技术。本项目将采用机会约束规划（risk-averse random programming, risk-averse random programming，规避不利随机性的一种重要变体）解决电力系统运行中的两个核心问题：实时调度和长期发电规划。研究目标和任务将：(1)应用基于抽样的方法建立机会约束的凸逼近；(2)开发分解方法并在并行计算框架中实现分解算法，以解决不确定大型电力系统的两个实例实际问题。开发的方法有望克服在可操作情况下面临的固有计算挑战。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 risk-averse decision-making models and methodologies to manage the integration of highly uncertain, large-scale renewable energy for society to benefit from cleaner, more reliable, and cost-effective power systems. The increase in renewable energy presents power systems with significant challenges due to the intermittency and limited predictability of production. Traditional operational strategies may be inadequate in the presence of such increased uncertainty when balancing power supply and demand economically and reliably. This CAREER project will bridge this pressing uncertainty gap to effectively integrate renewable energy and thus help to ensure system reliability and cost-effectiveness, which will bring transformative change to the power industry. The intellectual merits of the project include developing new risk-averse decision-making models to accommodate the nature of power systems and decision makers’ risk preferences, and developing new computationally efficient solution approaches to enable practical, useful solutions for all stakeholder and end-user communities. The broader impacts of the project include: (i) integrating research and education at the University of Arizona through mentoring undergraduate students in hands-on research for this CAREER project and connecting the research results with familiar applications to inspire the next generation of students to pursue a career in engineering; (ii) sustainable solutions to energy challenges in the Navajo Nation; and (iii) technology transfer to other applications in the presence of highly uncertainty for societally important industries in the US economy, such as transportation and finance.This project will adopt chance-constrained programming (an important variant of risk-averse stochastic programming against unfavorable randomness) to solve two core problems in power system operations: real-time dispatch and long-term generation planning. The research objectives and tasks will: (1) apply sampling-based methods to develop convex approximations for the chance constraints; and (2) develop decomposition approaches and implement the decomposition algorithms in a parallel-computing framework to solve the two example practical problems for uncertain, large-scale power systems. The developed approaches are expected to overcome the inherent computational challenges faced in actionable situations.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.