Hierarchical Predictive Control of an Unmanned Aerial Vehicle Integrated Power, Propulsion, and Thermal Management System

Increasing electrification of air vehicles presents challenges to the energy management of their coupled electric power, propulsion, and thermal systems. Notably, the complexity and multi-timescale nature of this class of systems make it difficult to design control algorithms that can simultaneously optimize dynamics across each energy domain. The unique contribution of this work is the design and experimental validation of a hierarchical model predictive controller that coordinates the electro-mechano-thermal dynamics of an unmanned aerial vehicle (UAV) integrated power, propulsion, and thermal management system. To support this contribution, a novel UAV energy system testbed is introduced. A graph-based framework is used to model the multi-domain dynamics of the UAV. To estimate the states of the experimental platform, a decentralized observer is described. When compared to a baseline approach, the hierarchical control strategy results in a higher performing and more reliable closed-loop system while decreasing fuel utilization by approximately 16%.