This thesis focuses on the modeling and control of a two-degree-of-freedom (2-DOF) helicopter system using the Quanser Aero 2 platform. Four different control strategies were developed and compared: two classical approaches (Proportional-Derivative, PD; and Linear Quadratic Regulator, LQR) and two nonlinear approaches (Sliding Mode Control, SMC; and Super-Twisting Control, STC), in order to address the challenges arising from the nonlinear and strongly coupled nature of the system dynamics.By deriving a nonlinear model based on the Lagrangian formalism and subsequently linearizing it, it was possible to implement linear controllers and design robust controllers to manage complex dynamics. The control strategies were validated through both MATLAB/Simulink simulations and real-time implementations using the QUARC software in combination with Arduino, with measurements obtained from pitch and yaw encoders.The experimental results confirmed the accuracy of the modeling and the effectiveness of the controllers, showing good correlation with simulations despite the inevitable approximations introduced by hardware, such as delays, noise, and saturations. The comparative analysis revealed that the linear controllers, PD and LQR, proved to be simple to implement and achieved satisfactory performance under standard operating conditions, while the nonlinear controllers, SMC and STC, although more complex to design, delivered superior performance in terms of robustness, response time, overshoot reduction, and reference tracking, both in simulation and in real-time implementation. In particular, the SMC controller demonstrated strong disturbance rejection capabilities, whereas the STC significantly reduced the chattering phenomenon typical of sliding mode control, thus achieving the best compromise between accuracy and stability. These results highlight how advanced control techniques, especially those based on sliding mode, represent effective solutions for managing complex dynamic systems such as the Quanser Aero 2. Future developments include the adoption of adaptive optimization algorithms, such as fuzzy logic, neural networks, and evolutionary methods, for automatic parameter tuning, as well as the introduction of artificial disturbances, for instance simulated wind, to further test the robustness of the designed controllers.
Questa tesi si concentra sulla modellazione e il controllo di un sistema elicotteristico a due gradi di libertà (2-DOF) utilizzando la piattaforma Quanser Aero 2. Sono state sviluppate e confrontate quattro diverse strategie di controllo: due classiche (Proporzionale-Derivativa, PD; e Regolatore Quadratico Lineare, LQR) e due non lineari (Controllo a Modalità Scorrevole, SMC; e Controllo Super-Twisting, STC), al fine di affrontare le sfide poste dalla natura non lineare e fortemente accoppiata della dinamica del sistema. Attraverso la derivazione di un modello non lineare basato sul formalismo lagrangiano e la sua successiva linearizzazione, è stato possibile implementare controllori lineari e progettare controllori robusti per la gestione di dinamiche complesse. Le strategie di controllo sono state validate sia tramite simulazioni MATLAB/Simulink sia tramite implementazioni in tempo reale utilizzando il software QUARC in combinazione con Arduino, con misure ottenute da encoder di beccheggio e imbardata. I risultati sperimentali hanno confermato l'accuratezza della modellazione e l'efficacia dei controllori, evidenziando una buona correlazione con le simulazioni, nonostante le inevitabili approssimazioni introdotte dall'hardware (ritardi, rumore, saturazioni). L'analisi comparativa ha mostrato che: I controllori lineari (PD e LQR) si sono dimostrati semplici da implementare, con prestazioni soddisfacenti in condizioni operative standard. I controllori non lineari (SMC e STC), sebbene più complessi da progettare, hanno fornito prestazioni superiori in termini di robustezza, tempo di risposta, riduzione dell'overshoot e inseguimento del riferimento, sia in simulazione che nell'implementazione reale. In particolare, il controllore SMC ha mostrato una buona capacità di mitigare l'effetto dei disturbi esterni, mentre l'STC ha ridotto significativamente il fenomeno di chattering tipico del controllo sliding mode, ottenendo il miglior compromesso tra accuratezza e stabilità. Questi risultati evidenziano come le tecniche di controllo avanzate, in particolare quelle basate sulla modalità scorrevole, rappresentino buone soluzioni per la gestione di sistemi dinamici complessi come il Quanser Aero 2. Le prospettive future includono l'uso di algoritmi di ottimizzazione adattiva (logica fuzzy, reti neurali, metodi evolutivi) per la regolazione automatica dei parametri e l'introduzione di disturbi artificiali (ad esempio vento simulato) per testare ulteriormente la robustezza dei controllori progettati.
Strategie di controllo lineari e non lineari per un twin rotor multivariabile
TRUFFELLI, IRENE
2024/2025
Abstract
This thesis focuses on the modeling and control of a two-degree-of-freedom (2-DOF) helicopter system using the Quanser Aero 2 platform. Four different control strategies were developed and compared: two classical approaches (Proportional-Derivative, PD; and Linear Quadratic Regulator, LQR) and two nonlinear approaches (Sliding Mode Control, SMC; and Super-Twisting Control, STC), in order to address the challenges arising from the nonlinear and strongly coupled nature of the system dynamics.By deriving a nonlinear model based on the Lagrangian formalism and subsequently linearizing it, it was possible to implement linear controllers and design robust controllers to manage complex dynamics. The control strategies were validated through both MATLAB/Simulink simulations and real-time implementations using the QUARC software in combination with Arduino, with measurements obtained from pitch and yaw encoders.The experimental results confirmed the accuracy of the modeling and the effectiveness of the controllers, showing good correlation with simulations despite the inevitable approximations introduced by hardware, such as delays, noise, and saturations. The comparative analysis revealed that the linear controllers, PD and LQR, proved to be simple to implement and achieved satisfactory performance under standard operating conditions, while the nonlinear controllers, SMC and STC, although more complex to design, delivered superior performance in terms of robustness, response time, overshoot reduction, and reference tracking, both in simulation and in real-time implementation. In particular, the SMC controller demonstrated strong disturbance rejection capabilities, whereas the STC significantly reduced the chattering phenomenon typical of sliding mode control, thus achieving the best compromise between accuracy and stability. These results highlight how advanced control techniques, especially those based on sliding mode, represent effective solutions for managing complex dynamic systems such as the Quanser Aero 2. Future developments include the adoption of adaptive optimization algorithms, such as fuzzy logic, neural networks, and evolutionary methods, for automatic parameter tuning, as well as the introduction of artificial disturbances, for instance simulated wind, to further test the robustness of the designed controllers.| File | Dimensione | Formato | |
|---|---|---|---|
|
ELABORATO_IRENE.TRUFFELLI.pdf
non disponibili
Dimensione
4.57 MB
Formato
Adobe PDF
|
4.57 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.12075/23682