Projecte llegit
Títol: Analysis and implementation of drone performance based trajectory estimation and prediction models
Estudiants que han llegit aquest projecte:
SÁNCHEZ SEGURA, ALBERT (data lectura: 23-07-2025)- Cerca aquest projecte a Bibliotècnica

Director/a: PASTOR LLORENS, ENRIC
Departament: DAC
Títol: Analysis and implementation of drone performance based trajectory estimation and prediction models
Data inici oferta: 11-02-2025 Data finalització oferta: 11-10-2025
Estudis d'assignació del projecte:
MU DRONS
Tipus: Individual | |
Lloc de realització: EETAC | |
Paraules clau: | |
Drone performance, u-space, strategic deconflicting, UTM, trajectory estimation, time prediction | |
Descripció del contingut i pla d'activitats: | |
Overview (resum en anglès): | |
Integrating drones into U-space airspace demands highly accurate and reliable performance modelling to effectively support strategic deconfliction and provide trustworthy 4D trajectory predictions. These capabilities are essential to ensure the safe, efficient and scalable integration of unmanned aircraft systems (UAS) into increasingly complex and dense airspace environments. As drone operations expand rapidly in both volume and variety, the need for accurate, adaptable, and robust models becomes critical to ensure that automated systems can coordinate traffic, predict conflicts, and mitigate risk effectively in real time. This significant challenge is being actively addressed in Europe through the ongoing development of the U-space framework, which aims to enable safe and efficient drone integration at scale.
However, most commercially available drones typically lack standardised performance profiles and well-defined operational values. This limitation makes current ATM-oriented models like BADA (Base of Aircraft Data), which are specifically tailored for general aviation, not directly usable for U-space services. These models rely on detailed manufacturer specifications and operational envelopes that are often unavailable or inconsistent for the diverse range of drone platforms, limiting their applicability. To address this gap, this paper presents a low computational, kinematic, operationally focused approach to drone performance modelling. The approach is based on classical dynamic principles and uses extensive real flight data collected from a diverse set of representative drone platforms operating in varying conditions. The methodology divides flight into key behavioural phases, such as turns, climbs, descents, cruise, and transitions. For each phase, polynomial estimators that capture essential dynamic behaviour without requiring complex subsystem modelling are generated. The proposed models are lightweight and packaged for easy integration by U-space service developers and validated against actual drone flight trajectories. The results demonstrate improved prediction accuracy, highlighting the value of tailored drone performance models in enabling safe, predictable and efficient U-space operations. |