CRII: CPS: Hierarchical Control for Constrained Multi-timescale Energy Management

现代控制系统，无论大小，都试图根据来自更多不同类型的传感器、传感模式、环境和环境的信息来优化资源。商用和国防飞机、车辆和船舶尤其如此，它们优化了燃油效率、性能和安全性。该研究项目开发了一种通用控制方法，以最大限度地提高整体性能和效率，即使在接近系统物理限制的情况下也是如此，同时确保安全并满足总体目标。本研究开发了一种新的“基于集合的方法”来构造具有可证明保证的网络物理系统(CPS)的有效递阶模型预测控制(MPC)，其中控制器之间的协调保证了执行器和状态约束的满足。该项目的研究描述了如何使用CPS系统约束为层次内的每个控制器构建特定的基于集合的协调机制，为整个系统提供可证明的闭环行为；如何利用系统的底层拓扑和动态时间尺度来开发特定于系统的分层控制体系结构，以最大限度地减少不确定性和最大化性能；以及如何将分层控制用于能源管理，例如，为电动汽车系统提供更高的能力、效率和可靠性。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。

Modern control systems, small and large, attempt to optimize resources based on information from more and different kinds of sensors, sensing modalities, environments, and circumstances. This is especially the case for commercial and defense aircraft, vehicles, and ships, that optimize for fuel efficiency, performance, and safety. This research project develops a general control methodology for maximizing overall performance and efficiency, even when operating close to the physical limits of the system, all while ensuring safety and meeting overall objectives. This research project develops a novel "set-based approach" to construct effective hierarchical model predictive control (MPC) with provable guarantees for cyber-physical systems (CPS), where coordination among controllers provides guaranteed satisfaction of actuator and state constraints. The research of this project describes how CPS system constraints can be used to construct specific set-based coordination mechanisms for each controller within the hierarchy that provide provable closed-loop behavior for the entire system, how the underlying topology and dynamic timescales of the system is leveraged to develop system-specific hierarchical control architectures that minimize uncertainty and maximize performance, and how hierarchical control can be used for energy management and, e.g., provide greater capability, efficiency, and reliability for electrified vehicle systems.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.