Design and Flight Testing of an Integral Linear Quadratic Regulator (LQRi) Attitude Control System for a Quadcopter UAV

Concetti Chiave

This paper presents the design, implementation, and flight test results of an integral linear quadratic regulator (LQRi) based attitude control system for a quadcopter UAV, which enhances the performance of traditional LQR by mitigating steady-state errors.

Sintesi

The paper begins by deriving the mathematical model for the kinematics and dynamics of the quadcopter UAV, including the linearized state-space representation about hover conditions. It then presents the design of both traditional LQR and LQRi controllers to stabilize the UAV in hover conditions and track desired attitude commands.
The controllers are implemented on the Pixhawk flight controller, and their performance is evaluated through high-fidelity software-in-the-loop (SITL) simulations and outdoor flight tests. The results show that the LQRi controller outperforms the traditional LQR, with a 30% improvement in attitude reference tracking in the SITL environment and a 40% improvement in roll and pitch axis tracking during the flight tests.
The paper also discusses the advantages of using the ArduPilot SITL environment for controller validation, including its high-fidelity simulation, seamless integration with the flight control software, and flexible configurability. Finally, the authors have made the code related to the LQRi controller implementation available for replication and further research.

Statistiche

The quadcopter used for flight testing had the following specifications:

Frame: DJI-F450
Motors: 920kv brushless
Propellers: 10 inch
Flight Controller: Pixhawk Cube Orange+
Power Source: 3s 4200mAh LiPo battery

Citazioni

"LQR with integral action, often referred to as Integral LQR (LQRi), enhances the standard LQR by incorporating integral control. This addition aims to eliminate steady-state errors and improve the system's response to modelling uncertainties or disturbances."
"The LQRi controller showed an improvement of over 30% against traditional LQR in attitude reference tracking."
"LQRi performed 40% better in trajectory tracking than traditional LQR in roll, and pitch axis, while being 13% better in the yaw axis."

Approfondimenti chiave tratti da

Design And Flight Testing Of LQRi Attitude Control For Quadcopter UAV

by Astik Srivas... alle arxiv.org 04-19-2024

https://arxiv.org/pdf/2404.12261.pdf

Domande più approfondite

How can the LQRi controller be further optimized to improve its performance, especially in the yaw axis?

To enhance the performance of the LQRi controller, particularly in the yaw axis, several optimization strategies can be implemented. One approach is to fine-tune the weighting matrices Q and R in the cost function of the LQRi controller. By adjusting these matrices, the controller's sensitivity to errors and control effort can be optimized, leading to improved tracking accuracy in all axes, including yaw. Additionally, implementing an adaptive tuning mechanism that dynamically adjusts these matrices based on real-time performance feedback can further enhance the controller's robustness and adaptability to varying conditions.
Furthermore, integrating a feedforward control component into the LQRi framework can help compensate for disturbances and external factors affecting the quadcopter's yaw dynamics. By incorporating predictive elements that anticipate disturbances and proactively adjust control inputs, the controller can preemptively counteract external influences, resulting in smoother and more precise yaw control. Additionally, implementing a state observer or estimator to improve state estimation accuracy can enhance the controller's ability to predict and respond to yaw deviations effectively.

What other control strategies, such as adaptive or robust control, could be combined with the LQRi approach to enhance the quadcopter's ability to handle uncertainties and disturbances?

Combining the LQRi approach with adaptive control techniques can significantly improve the quadcopter's resilience to uncertainties and disturbances. Adaptive control mechanisms, such as Model Reference Adaptive Control (MRAC) or Adaptive Sliding Mode Control, can continuously adjust controller parameters based on real-time system identification, enabling the quadcopter to adapt to changing environmental conditions and disturbances. By integrating adaptive elements into the LQRi framework, the controller can autonomously optimize its performance and maintain stability in the presence of varying dynamics.
Moreover, incorporating robust control strategies, such as H-infinity control or μ-synthesis, alongside the LQRi approach can further enhance the quadcopter's ability to handle uncertainties and disturbances. Robust control techniques focus on minimizing the impact of uncertainties in the system by designing controllers that are inherently stable and resilient to variations in the system parameters. By combining robust control methods with LQRi, the quadcopter can achieve a higher level of stability and performance across a wide range of operating conditions, making it more reliable in challenging environments.

What potential applications or scenarios could benefit the most from the improved attitude control accuracy provided by the LQRi controller, and how could it be leveraged to enable new capabilities for quadcopter UAVs?

The enhanced attitude control accuracy offered by the LQRi controller opens up a range of potential applications and scenarios where precise maneuverability and stability are crucial. One key area that could benefit significantly is aerial photography and videography, where the quadcopter's ability to maintain a steady orientation and track specific targets is essential for capturing high-quality footage. By leveraging the improved control accuracy of the LQRi controller, quadcopters can achieve smoother and more controlled movements, resulting in superior aerial imaging capabilities.
Additionally, applications in search and rescue operations could greatly benefit from the enhanced attitude control provided by the LQRi controller. In scenarios where quadcopters are deployed to locate and assist individuals in emergency situations, the ability to navigate accurately and maintain stable flight is critical. The precise attitude control offered by the LQRi controller can enable quadcopters to maneuver through complex environments with greater agility and responsiveness, improving their effectiveness in search and rescue missions.
Furthermore, the improved control accuracy of the LQRi controller can enable new capabilities for quadcopter UAVs in autonomous navigation and obstacle avoidance. By leveraging the controller's enhanced tracking performance, quadcopters can autonomously follow predefined paths, avoid obstacles in real-time, and execute complex flight maneuvers with precision. This opens up possibilities for applications in surveillance, inspection, and monitoring tasks where autonomous operation and accurate control are paramount.