Robust adaptive control has received much attention nowadays thanks to its capability to ensure satisfactory performance level in systems characterized by nonlinear dynamics and a high level of uncertainty, which is often the case for high-performance with aerospace systems that operate in a large domain of conditions. This thesis focuses on the development of a position controller for the quadrotor Unmanned Aerial Vehicles (UAVs) affected by uncertainties in the dynamic system, such as unmodeled aerodynamic drags, using adaptive control strategies. In the first part of this thesis the current State of the Art regarding Adaptive Control is studied, then, the dynamics of the quadrotor is described with focus on the aerodynamic effects that may cause instability or deterioration of performance at high speed flight. The second part of the thesis presents the design process for the position controller, specifically the augmentation approach is adopted in which the adaptive laws are added on top of a baseline controller designed on the nominal plant, i.e., the open-loop dynamic system free of uncertainties. The strategies used for adaptive augmentation are the Model Reference Adaptive Control (MRAC) and the L1 Adaptive Control, with the derivation of an ad hoc uncertainties parameterization including the second order drag term. For the L1 Adaptive strategy, in addition to the standard implementation, a different architecture that uses only the control output of the baseline controller is implemented for the quadrotor. Also a modified version of the baseline controller that includes a feedforward term to compensate aerodynamic drags is implemented in order to compare with adaptive strategies. The benefits of the augmentation are evaluated first in a simulator in Matlab/Simulink with the quadrotor dynamic model implemented, then flight test experiments are carried out in the Flying Arena for Rotorcraft Technologies (FlyART) of Aerospace Systems and Control Laboratory (ASCL) of Politecnico di Milano by comparing the performance of baseline and augmented controllers in real case scenarios.
Il controllo adattivo robusto ha ricevuto molta attenzione al giorno d’oggi grazie alla sua capacità di garantire un livello di prestazione soddisfacente in sistemi caratterizzati da dinamica non lineare e un alto livello d’incertezza, come spesso accade per sistemi aerospaziali ad alte prestazioni che operano in un ampio dominio di condizioni. Questa tesi si concentra sullo sviluppo di un controllore di posizione per i qudrotors (Unmanned Aerial Vehicles UAV) affetti da incertezze nel sistema dinamico, ad esempio le resistenze aerodinamiche non modellate, utilizzando strategie di controllo adattivo. Nella prima parte di questa tesi si studia l’attuale Stato dell’Arte in merito al Controllo Adattivo, poi si descrive la dinamica del quadrotor focalizzandosi sugli effetti aerodinamici che possono causare instabilità o deterioramento delle prestazioni in regime ad alta velocità. La seconda parte della tesi presenta il processo di progettazione per il controllore di posizione, in particolare viene adottato l’approccio dell’aumento in cui le leggi adattative vengono aggiunte sopra un controllore di base (baseline controller) progettato sul sistema nominale, i.e., il sistema dinamico ad anello aperto privo di incertezze. Le strategie utilizzate per l’aumento adattivo sono il Model Reference Adaptive Control (MRAC) e il L1 Adaptive Control, con la derivazione di una parametrizzazione ad hoc delle incertezze che include il termine aerodinamico del secondo ordine. Per la strategia L1 Adaptive, oltre all’implementazione standard per il quadrotor, viene implementata un’architettura diversa che utilizza solo il control output del baseline controller. Viene implementata anche una versione modificata del baseline controller che include un termine in feedforward per compensare le resistenze aerodinamiche al fine di confrontare con le strategie adattive. I vantaggi delle strategie adattative vengono prima valutati in un simulatore in Matlab/Simulink con il modello dinamico del quadrotor implementato, sucessivamente vengono effettuati esperimenti di volo nella Flying Arena per Rotorcraft Technologies (FlyART) di Aerospace Systems and Control Laboratory (ASCL) del Politecnico di Milano confrontando la performance del baseline controller e le strategie adattative.
Adaptive control for high-performance trajectory tracking in quadrotor UAVs
LIU, SHANG
2021/2022
Abstract
Robust adaptive control has received much attention nowadays thanks to its capability to ensure satisfactory performance level in systems characterized by nonlinear dynamics and a high level of uncertainty, which is often the case for high-performance with aerospace systems that operate in a large domain of conditions. This thesis focuses on the development of a position controller for the quadrotor Unmanned Aerial Vehicles (UAVs) affected by uncertainties in the dynamic system, such as unmodeled aerodynamic drags, using adaptive control strategies. In the first part of this thesis the current State of the Art regarding Adaptive Control is studied, then, the dynamics of the quadrotor is described with focus on the aerodynamic effects that may cause instability or deterioration of performance at high speed flight. The second part of the thesis presents the design process for the position controller, specifically the augmentation approach is adopted in which the adaptive laws are added on top of a baseline controller designed on the nominal plant, i.e., the open-loop dynamic system free of uncertainties. The strategies used for adaptive augmentation are the Model Reference Adaptive Control (MRAC) and the L1 Adaptive Control, with the derivation of an ad hoc uncertainties parameterization including the second order drag term. For the L1 Adaptive strategy, in addition to the standard implementation, a different architecture that uses only the control output of the baseline controller is implemented for the quadrotor. Also a modified version of the baseline controller that includes a feedforward term to compensate aerodynamic drags is implemented in order to compare with adaptive strategies. The benefits of the augmentation are evaluated first in a simulator in Matlab/Simulink with the quadrotor dynamic model implemented, then flight test experiments are carried out in the Flying Arena for Rotorcraft Technologies (FlyART) of Aerospace Systems and Control Laboratory (ASCL) of Politecnico di Milano by comparing the performance of baseline and augmented controllers in real case scenarios.File | Dimensione | Formato | |
---|---|---|---|
Excutive_summary_SHANG_LIU.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: Excutive_summary
Dimensione
1.15 MB
Formato
Adobe PDF
|
1.15 MB | Adobe PDF | Visualizza/Apri |
THESIS_SHANG_LIU.pdf
accessibile in internet solo dagli utenti autorizzati
Descrizione: THESIS
Dimensione
8.29 MB
Formato
Adobe PDF
|
8.29 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/10589/196909