Abstract
dc:description.abstractAutonomous flying vehicles are becoming more and more prominent nowadays and have manifold applications, some only about to emerge: collecting traffic data, surveillance and security, disaster management, wildlife observation, delivery services, and defence. Providing such vehicles with enough computational intelligence to automatically steer according to given objectives is a challenging task, only recently tackled with control algorithms working in soft-real time and computationally feasible for the limited on-board resources. Researchers and practitioners historically address the problem by separately concentrating on the two sub-problems of vehicle routing (VR) and trajectory optimisation (TO), often neglecting or oversimplifying vehicles dynamic. In this work we introduce methodologies contributing to the newborn research area unifying the two aspects in a single Vehicle Routing and Trajectory Optimisation problem (VRTOP), also taking into account a sufficiently accurate representation of vehicles dynamic and providing solution algorithms to the overall problem. First, a model for the VRTOP is proposed where trajectory constraints are linearised through a Taylor expansion around a known solution to vehicles’ equations of motion within predefined settings. Distance constraints are also linearised to arrive at a Mixed-Integer Linear Programming formulation of the VRTOP with prior information. Second, this work provides approximation techniques to gain valuable insights and find initial trajectory guesses when prior knowledge is not available, overcoming the local nature of Taylor’s approximation and resulting in enhanced and more robust solutions. Finally, a Two Step Solution Algorithm is proposed to integrate the two phase of the problem, enabling us to solve the VRTOP in absence of prior information. Numerical results showing the efficacy of the presented approach are furnished and further research developments are ultimately discussed.
Degree
thesis:*- Name dc:type.qualificationname
- Ph.D.
- Level dc:type.qualificationlevel
- doctoral
- Grantor dc:publisher.institution
- University of Southampton
- Year dc:date.issued
- 2024
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Bobbio, Luigi
- Advisors dc:contributor.advisor
-
- Fliege, Joerg
- Martinez Sykora, Toni
- ORCID person:identifier.orcid
Rights
- Language dc:language
- English, eng