COMPARATIVE ANALYSIS OF CLOSED LOOP TUNING METHODS FOR PID CONTROLLER ON PX4 QUADCOPTER SYSTEM

• Main Author: Juliansah, Joni
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/86853
• Institution: Institut Teknologi Bandung

A quadcopter is a type of UAV (Unmanned Aerial Vehicle) characterized by complex nonlinear dynamics, requiring a reliable control system to ensure its stability and responsiveness. This study aims to evaluate the performance of PID (Proportional-Integral-Derivative) control on the X, Y, and Z (altitude) axes of a quadcopter using closed-loop tuning methods such as Ziegler-Nichols, Cohen- Coon, Tyreus-Luyben, Åström-Hägglund, and IMC-Based PID. The research utilizes the Quadcopter with MATLAB PX4 Support Hardware Toolbox, Simulink, and the jMAVsim simulator, integrating PID control for various tuning methods. Tuning parameters were determined through critical oscillation tests to derive the optimal values for Kp, Ki, and Kd. The performance of each method was compared based on four evaluation criteria: rise time, settling time, overshoot, and steady- state error. The simulation results indicate significant differences among the PID tuning methods. Åström-Hägglund demonstrated the best overall performance with fast rise time, short settling time, low overshoot, and minimal steady-state error. Cohen- Coon achieved a balance between quick response and stability, although it still exhibited a small steady-state error. IMC-Based PID excelled in achieving the fastest rise time but had a larger steady-state error compared to other methods. Ziegler-Nichols provided a quick response but suffered from high overshoot and long settling time, making it less suitable for dynamic systems like quadcopters. Tyreus-Luyben had the poorest performance, with slow rise time, high overshoot, and very long settling time. The conclusion of this study indicates that Åström-Hägglund is the most optimal method for position control of a quadcopter on the X, Y, and Z axes, as it delivers fast, stable, and accurate responses. Recommendations for future development include implementing the method on a physical quadcopter system, optimizing parameters using heuristic methods, and testing under external disturbance conditions to enhance system adaptability.