This paper presents a comprehensive framework for energy-efficient trajectory planning and feasibility assessment in drone package delivery operations across urban areas. The framework is designed to assist unmanned aerial systems operators within the Unmanned Aircraft System Traffic Management architecture by generating feasible operational plans for a given set of package delivery missions. The framework addresses the multifaceted computational challenges of this problem using a two-layer approach. The upper layer consists of a Markov-decision-process-based trajectory planner that generates energy-efficient trajectories by considering factors such as the aircraft model, the length and smoothness of the trajectories, and the interaction of the aircraft with the wind field. The lower layer then assesses the feasibility of these energy-efficient trajectories through a battery state of charge prediction-based uncertainty quantification scheme. We demonstrate the efficacy of the proposed framework in a realistic drone package delivery scenario set in an urban environment in Boston City. In addition, an actual flight experiment was conducted to validate the framework’s performance. The results highlight the framework’s capability to generate energy-efficient trajectories and assess the feasibility of each mission. Source code: github.com/Abenezergirma/ETP-FA-Package-Delivery.
Energy-Efficient Trajectory Planning and Feasibility Assessment Framework for Drone Package Delivery
2025-03-01
Conference paper , Article (Journal)
Electronic Resource
English