CPS: Breakthrough: Collaborative Research: Transactive Control of Smart Railway Grid

该项目采用智能网络物理方法，以改善美国和其他国家的电气化铁路基础设施。我们将在确保整个铁路基础设施的端到端时间表限制的同时，开发定价和能源利用的分布式协调。我们将通过交互控制来确保这种分布式协调，交互控制是一种明智的网络物理系统动态定价设计，该系统利用计算和通信基础设施，并适应基础列车服务的物理限制。该项目具有协同作用，因为它建立在UIC的电动火车基础设施和协调的专业知识以及MIT的交易控制的基础上。 我们将通过与宾夕法尼亚州东南交通管理局的工程师合作来验证这种方法，该管理局正在进行重大的现代化工作，以改善其电动火车系统。该项目还涉及强有力的国际合作，这也将使技术的验证。该项目将制定一个多尺度的传递控制策略，以最大限度地降低地理上分散的铁路网的价格和能源利用率，并对不断发展的智能和微电网产生更广泛的影响。这些交易在不同的时间尺度上发展，从电网和铁路系统运营商之间的日前离线交易产生价格最优到列车或区域控制中心（ACC）之间的实时最优交易。所有这些交易都是在满足系统约束的同时进行的，这些约束包括端到端的时间调度、电能质量和容量。我们的研究重点是基本问题，包括信息，控制和电源的集成，包括事件驱动的数据包从源节点到达目的地节点，同时确保硬相对的最后期限和最佳采样和传感;和制定网络凹效用函数分配有限的通信网络容量之间的控制回路。 该项目开发的优化方法可以类似地应用于多个应用领域。

This project pursues a smart cyber-physical approach for improving the electric rail infrastructure in the United States and other nations. We will develop a distributed coordination of pricing and energy utilization even while ensuring end-to-end time schedule constraints for the overall rail infrastructure. We will ensure this distributed coordination through transactive control, a judicious design of dynamic pricing in a cyber-physical system that utilizes the computational and communication infrastructure and accommodates the physical constraints of the underlying train service. The project is synergistic in that it builds upon the expertise of the electric-train infrastructure and coordination at UIC and that of transactive control on the part of MIT. We will validate the approach through collaboration with engineers in the Southeastern Pennsylvania Transport Authority, where significant modernization efforts are underway to improve their electric-train system. The project also involves strong international collaboration which will also enable validation of the technologies.This project will formulate a multi-scale transitive control strategy for minimization of price and energy utilization in a geographically-dispersed railway grid with broader implications for evolving smart and micro grids. The transactions evolve over different temporal scales ranging from day-ahead offline transaction between the power grid and the railway system operators yielding price optimality to real-time optimal transaction among the trains or the area control centers (ACC). All of these transactions are carried out while meeting system constraints ranging from end-to-end time-scheduling, power-quality, and capacity. Our research focuses on fundamental issues encompassing integration of information, control, and power, including event-driven packet arrival from source to destination nodes while ensuring hard relative deadlines and optimal sampling and sensing; and formulation of network concave utility function for allocating finite communication-network capacity among control loops. The project develops optimization approaches that can be similarly applied across multiple application domains.