Optimal Task Scheduling and Flight Planning for Multi-Task Unmanned Aerial Vehicles
- Publication Type:
- Thesis
- Issue Date:
- 2022
Open Access
Copyright Clearance Process
- Recently Added
- In Progress
- Open Access
This item is open access.
Unmanned aerial vehicles (UAVs), also known as drones, play an important role in various areas due to their agility and versatility. Integrated with many embedded components, the UAV is capable of conducting multiple tasks simultaneously. Coordinating different tasks to a multi-task UAV can be challenging. The reason is that tasks may require different levels of commitment and tolerate different latencies. Another reason is that multi-tasking can give rise to difficulties in the UAV’s energy management, as many UAVs are battery-powered. In this thesis, we study the optimal flight planning, control, and routing for the multi-task UAV. The main contributions of this thesis can be summarized as follows.
• This thesis presents a novel energy-efficient UAV flight planning framework, which integrates UAVs into intelligent transportation systems for energy-efficient, delay-sensitive delivery services. The UAV can dynamically choose actions from cruise speed, full speed, recharging at a roadside charging station, or hitchhiking and recharging on a collaborative vehicle. The objective is to minimize the energy consumption of the UAV and ensure timely delivery. We reveal the conditions under which the UAV’s flight planning changes in terms of the remaining flight distance or the elapsed time. Consequently, the optimal flight planning can be instantly made by comparing with the thresholds.
• This thesis presents a new online control framework for multi-task UAVs, which allows a UAV to perform in-situ sensing while delivering goods. A new finite-horizon Markov decision process (FH-MDP) problem is formulated to ensure timely delivery, minimize the UAV’s energy consumption, and maximize its reward for in-situ sensing. We prove the monotonicity and subadditivity of the FH-MDP, such that the FH-MDP has an optimal, monotone deterministic Markovian policy. We find that the optimal policy consists of flight distance-related and time-related thresholds at which the optimal action of the UAV switches. As a result, the optimal actions of the UAV can be obtained by comparing its state with the thresholds at a linear complexity.
• This thesis presents a novel multi-task UAV routing framework, which aims to minimize the UAV’s energy consumption, maximize its sensing reward, and ensure its timely arrival at the destination. We interpret possible flight waypoints as location-dependent tasks, hence accommodating the waypoints and in-situ sensing in a unified process of task selection. We construct a weighted time-task graph, and transform the optimal routing of the UAV into a weighted routing problem, which can be optimally solved by the celebrated Bellman-Ford algorithm.
Please use this identifier to cite or link to this item: