Hierarchical control for multi-domain coordination of vehicle energy systems with switched dynamics

Dynamic modeling;Physics-based modeling;Graph-based model;Control;Hierarchical control;Model Predictive Control;Passivity;Monotone system;Cooperative system;Energy management;Vehicle energy management;Thermal management;Electro-thermal system;Vehicle system;Aircraft energy system

Pangborn, Herschel C.

关键词: Dynamic modeling;    Physics-based modeling;    Graph-based model;    Control;    Hierarchical control;    Model Predictive Control;    Passivity;    Monotone system;    Cooperative system;    Energy management;    Vehicle energy management;    Thermal management;    Electro-thermal system;    Vehicle system;    Aircraft energy system;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/104799/PANGBORN-DISSERTATION-2019.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

This dissertation presents a hierarchical control framework for vehicle energy management.As a result of increasing electrification, legacy integration and control approaches for vehicle energy systems have become limiting factors of performance and cannot accommodate the requirements of next-generation systems. Addressing this requires control frameworks that coordinate dynamics across multiple physical domains and timescales, enabling transformative improvements in capability, efficiency, and safety. To capture multi-domain storage and exchange of energy, a graph-based dynamic modeling approach is proposed and experimentally validated. This modeling approach is then leveraged for model-based control, in which the complex task of energy management is decomposed into a hierarchical network of model predictive controllers that coordinate decision-making across subsystems, physical domains, and timescales. The controllers govern both continuous and switched dynamic behaviors, addressing the hybrid nature of modern vehicle energy systems.The proposed hierarchical control framework is evaluated in application to a hardware-in-the-loop electro-thermal testbed representative of a scaled aircraft energy system, where it achieves significantly improved capability, efficiency, and safety as compared to legacy control approaches. Next, the structural information embedded in the graph-based modeling approach is shown to facilitate analysis. Closed-loop stability of decentralized MPC frameworks is guaranteed by analyzing the passivity of switched nonlinear graph-based systems and augmenting their controllers with a local passivity-based constraint. Lastly, a hierarchical control formulation guaranteeing satisfaction of state and input constraints for a class of switched graph-based systems is presented. This formulation is demonstrated in application to thermal management using both simulation and experimental implementation.