In the context of this thesis, a Model Predictive Control algorithm was implemented to control a nano-quadcopter swarm. Several experimental tests were performed with up to four drones and without using a tracking system, successfully reaching different goals as the achievement of targets imposed to the center of mass of the swarm or to each drone, the avoidance of obstacles known a priori, the prevention of collisions between drones, and coordinated flight. The high computational effort and the error state estimation were investigated. Crazyflie 2.1 drones equipped with V2 flow deck were used for performing experimental tests, whereas the MPC algorithm was managed by a ground-based computer able to communicate with drones via radio.
Nel contesto di questa tesi, è stato implementato un algoritmo Model Predictive Control per controllare uno sciame di nano-quadricotteri. Sono stati eseguiti diversi test sperimentali con un massimo di quattro droni e senza l'utilizzo di un sistema di tracciamento, raggiungendo con successo diversi obiettivi come il raggiungimento di target imposti a livello centro di massa dello sciame o per ciascun drone, l'evitamento di ostacoli noti a priori, la prevenzione di collisioni tra droni e volo coordinato. L'elevato costo computazionale e l'errore nella stima dello stato sono stati analizzati. Per l'esecuzione dei test sperimentali sono stati utilizzati droni Crazyflie 2.1 dotati di flow deck V2, mentre l'algoritmo MPC è stato gestito da un computer a terra in grado di comunicare con i droni via radio.
Model Predictive Control applied to a Nano-Quadcopter Swarm
Viel, Matteo
2022/2023
Abstract
In the context of this thesis, a Model Predictive Control algorithm was implemented to control a nano-quadcopter swarm. Several experimental tests were performed with up to four drones and without using a tracking system, successfully reaching different goals as the achievement of targets imposed to the center of mass of the swarm or to each drone, the avoidance of obstacles known a priori, the prevention of collisions between drones, and coordinated flight. The high computational effort and the error state estimation were investigated. Crazyflie 2.1 drones equipped with V2 flow deck were used for performing experimental tests, whereas the MPC algorithm was managed by a ground-based computer able to communicate with drones via radio.File | Dimensione | Formato | |
---|---|---|---|
2023_07_Viel_01.pdf
non accessibile
Dimensione
12.25 MB
Formato
Adobe PDF
|
12.25 MB | Adobe PDF | Visualizza/Apri |
2023_07_Viel_02.pdf
non accessibile
Dimensione
1.46 MB
Formato
Adobe PDF
|
1.46 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/210354