Deep-space missions have so far consisted of large, heavy, and expensive spacecraft operated by engineers from the Earth. In the last two decades, a new platform paradigm has begun to be employed for Earth-based space missions. These new platforms are widely known as CubeSats, shoebox-sized spacecraft that drastically reduce the development and integration costs of classical probes. CubeSats have recently started to be considered for deep-space missions as well. Despite their reduced costs, they are still operated by engineers from the Earth, meaning that their operational costs are comparable to those of conventional spacecraft. Moreover, since the number of CubeSats launches is expected to rapidly increase, ground-station slots will saturate soon. To reduce the operational costs and prevent the saturation of ground station slots, spacecraft with autonomous capabilities are therefore desired, if not required. The ERC-funded EXTREMA project collocates in this context, and this work is part of the project. In particular, this dissertation focuses on autonomous guidance of deep-space CubeSats. The low-thrust trajectory optimization problem is first analyzed from a theoretical perspective. Convex optimization is then exploited to build a reliable and rapid algorithm that solves the low-thrust trajectory optimization problem. Operational aspects are subsequently considered, such as the need for an exact bang-bang thrust profile to be executed onboard. A guidance algorithm architecture is proposed and tested on space-like hardware in terms of computational time and required memory, and its output is also executed by a hardware facility. An integrated guidance and navigation algorithm is proposed. The developed algorithm turns out to be a tool that can be used for preliminary mission analysis purposes. In particular, this dissertation presents the porkchop plots for a low-thrust option of ESA's RAMSES mission to asteroid Apophis, which will close-encounter the Earth in April 2029. Finally, this dissertation introduces an approach that models uncertainties on the thruster for stochastic low-thrust trajectory optimization.
Fino ad oggi, le missioni spaziali interplanetarie hanno utilizzato satelliti di grandi dimensioni, pesanti, e soprattutto costosi, operati da ingegneri/e sulla Terra. Nel corso degli ultimi due decenni, un nuovo tipo di piattaforma ha iniziato ad essere utilizzato per missioni attorno al nostro pianeta. Questo nuovo tipo di satelliti è meglio conosciuto come CubeSats, la cui forma e dimensioni ricordano quelle di una scatola di scarpe. Essi hanno permesso di ridurre i costi di sviluppo e integrazione delle piattaforme spaziali. I CubeSats sono stati recentemente considerati anche per missioni nello spazio profondo. Tuttavia, essi sono ancora controllati dagli/dalle ingegneri/e sulla Terra, e quindi il costo delle loro operazioni è pari a quello dei satelliti tradizionali. In aggiunta, il numero di lanci di CubeSats è destinato a crescere notevolmente e per questo gli slots di comunicazione con le antenne di Terra satureranno presto. Per risolvere questi problemi, satelliti con capacità autonome sono dunque desiderabili o addirittura necessari. Il progetto EXTREMA, finanziato dal Consiglio Europeo della Ricerca, si colloca in questo contesto, e questa tesi si focalizza sulla guida autonoma dei CubeSats nello spazio profondo. Il problema di ottimizzare le traiettorie spaziali è dapprima analizzato sotto un aspetto teorico. Successivamente, l'ottimizzazione convessa viene utilizzata per definire un algoritmo che risolve il problema in maniera rapida e affidabile. Aspetti operativi sono poi considerati, come la necessità di un profilo di spinta esattamente bang-bang. La struttura di un algoritmo di guida autonoma è proposto e testato in termini di memoria richiesta e tempo computazionale su processori simili a quelli usati nello spazio. L'algoritmo sviluppato è anche un ottimo strumento per effettuare analisi di missione preliminari. La tesi presenta i porkchop plots relativi alla missione RAMSES dell'Agenzia Spaziale Europea verso l'asteroide Apophis, che effettuerà un incontro ravvicinato con la Terra nell'aprile 2029. In conclusione, questa tesi introduce un approccio per modellare le incertezze dovute al motore che permette di risolvere un problema di controllo ottimo stocastico.
Convex optimal control for on-board and on-ground low-thrust spacecraft guidance
Morelli, Andrea Carlo
2023/2024
Abstract
Deep-space missions have so far consisted of large, heavy, and expensive spacecraft operated by engineers from the Earth. In the last two decades, a new platform paradigm has begun to be employed for Earth-based space missions. These new platforms are widely known as CubeSats, shoebox-sized spacecraft that drastically reduce the development and integration costs of classical probes. CubeSats have recently started to be considered for deep-space missions as well. Despite their reduced costs, they are still operated by engineers from the Earth, meaning that their operational costs are comparable to those of conventional spacecraft. Moreover, since the number of CubeSats launches is expected to rapidly increase, ground-station slots will saturate soon. To reduce the operational costs and prevent the saturation of ground station slots, spacecraft with autonomous capabilities are therefore desired, if not required. The ERC-funded EXTREMA project collocates in this context, and this work is part of the project. In particular, this dissertation focuses on autonomous guidance of deep-space CubeSats. The low-thrust trajectory optimization problem is first analyzed from a theoretical perspective. Convex optimization is then exploited to build a reliable and rapid algorithm that solves the low-thrust trajectory optimization problem. Operational aspects are subsequently considered, such as the need for an exact bang-bang thrust profile to be executed onboard. A guidance algorithm architecture is proposed and tested on space-like hardware in terms of computational time and required memory, and its output is also executed by a hardware facility. An integrated guidance and navigation algorithm is proposed. The developed algorithm turns out to be a tool that can be used for preliminary mission analysis purposes. In particular, this dissertation presents the porkchop plots for a low-thrust option of ESA's RAMSES mission to asteroid Apophis, which will close-encounter the Earth in April 2029. Finally, this dissertation introduces an approach that models uncertainties on the thruster for stochastic low-thrust trajectory optimization.| File | Dimensione | Formato | |
|---|---|---|---|
|
main.pdf
accessibile in internet per tutti
Descrizione: PhD Thesis
Dimensione
14.54 MB
Formato
Adobe PDF
|
14.54 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/10589/220732