Apply on
Availability Status
This job is expected to be in high demand and may close soon. We’ll remove this job ad once it's closed.
INTERNSHIP TOPIC: Generation of drone/swarm trajectories for Radar evaluation
Description:
The emergence of illegal, malicious and dangerous swarm drones create new challenges for Radar systems. The presence of a high number of closed objects represents a complex tracking challenge. This complexity arises from several factors: measurements generated by closed-spaced targets lead to difficulties to form stable state estimates and trajectories for each target, poor tracking accuracy (accuracy of predicted states, measurement errors) and explosion of combinations in association algorithms making data association much less valid.
In this internship, it is proposed to study the generation of drone swarm and flock of birds' trajectories (including flight mechanics and swarm model dynamics) for the evaluation of the tracking behavior using a Radar Digital Twin. This will help identify future algorithms for tracking optimisation.
- The study will start with research of the data available for drone and birds' trajectories (both single object and swarms of objects) and an analysis of the different models of swarm dynamics: boids, vicsek, etc…
- The next step will be to define swarm scenarios and generate corresponding trajectories using simulators for drone/swarm developed by ENAC.
- Once the swarm trajectories will be generated, they will be played on the Radar Digital Twin and the behaviour of the tracker with respect to swarm of objects scenarios could be characterised.
Specific Responsibilities:
Research of the data available, analysis of swarm dynamics models
Study of drone/swarm simulators
Generation of trajectories and data format conversion for the Radar Digital Twin Run trajectory scenarios and characterization of the tracking behaviour
Key Requirements:
Keen interest in flight mechanics and radar
Skills in scripting/programming (Python)
Knowledge in data processing would be advantageous
Expected Outcome:
Drone/swarm trajectories
Characterization of the tracking behaviour