在本研究中，探討了模型參考適應性PID控制器（MRAC-PID）在四旋翼無人機上的應用和性能。此研究目的是優化傳統PID控制器，特別是對抗外部干擾與不同系統類型的情況下，而首先會先建立了無人機的數學模型。
在控制器設計方面，本篇將傳統的PID控制方法與適應性調整機制結合，利用Matlab Simulink模擬來評估控制器對連續干擾、高斯雜訊以及突波等外部信號干擾的反應，並利用標準二階系統的響應當作參考模型與四旋翼無人機模型做對照來調整系統參數。
經過模擬之數據比較後，適應性PID控制器相比於傳統PID控制器，適應性PID控制器在各種干擾情形下最多能夠減少平均誤差量42%、平均最大超越量則能有效的減少最多96%，透過模擬與數據分析可以證明適應性控制器能夠優化傳統PID控制。
本研究中透過嵌入式系統實時修改飛行參數，並利用一系列的實驗來測試在模擬階段所提出的方法，而實驗結果顯現，適應性控制器在實際無人機系統中能夠減少平均誤差量50%，而平均最大超越量也能夠減少62%，也透過實驗確保理論與應用之間的關聯性與實用性。

In this study, the application and performance of Model Reference Adaptive PID Controllers (MRAC-PID) on quadrotor drones is explored. The main objective of this research is to optimize the performance of traditional PID controllers, particularly in combating external disturbances and across different system types. Initially, a mathematical model of the drone and developed a control strategy based on MRAC is established, enabling the control system to automatically adjust the PID parameters in response to environmental changes and system dynamics.
In terms of controller design, traditional PID control methods with adaptive adjustment mechanisms is conbined, utilizing Matlab Simulink simulations to assess the controller's response to continuous disturbances, Gaussian noise, and impulse signal interferences. It is also used the response of a standard second-order system as a reference model to adjust the system parameters of the quadrotor model, ensuring that the adaptive controller could achieve ideal response outputs under various operational conditions.
Simulation results are shown that the adaptive PID controller has a better response in terms of disturbance rejection and response speed compared to traditional PID controllers. Through dynamic adjustment of PID parameters, the adaptive controller is more effective in handling external disturbances, maintaining system stability and performance. Additionally, it is investigated the impact of the learning rate on control performance, finding that the appropriate setting of the learning rate is crucial for achieving optimal control effects.
By interfacing with embedded systems and various libraries to modify flight parameters, and by validating the simulation results through practical experiments, this study demonstrates the effectiveness and feasibility of model reference adaptive PID controllers in the application of quadrotor drones, and provides valuable insights for the development and optimization of future drone control systems.

[1] Bora Erginer and Erdinç Altuğ , “Modeling and PD Control of a Quadrotor VTOL Vehicle” ,Proceedings of the 2007 IEEE Intelligent Vehicles Symposium Istanbul, Turkey, June 13-15 (2007)
[2] Jun Li and Yuntang Li ,“Dynamic Analysis and PID Control for a Quadrotor,” IEEE International Conference on Mechatronics and Automation, Beijing, China, 2011, pp.　573-578, doi: 10.1109/ICMA.2011.5985724. (2011)
[3] Ioan Doré Landau, Rogelio Lozano, Mohammed M’Saad, Alireza Karimi, “Adaptive Control Algoritms, Analysis and Applications - Second Edition”, ISBN:978-0-85729-663-4 (2011)
[4] Anton Glushchenko, Konstantin LastochKin, “Neural Network-based Direct Model Reference Adaptive Control of Quadrotor Attitude”, 978-1-6654-6586-1 ©2022 IEEE (2022)
[5] Matthias Schreier, “Modeling and Adaptive Control of a Quadrotor” DOI: 10.1109/ICMA.2012.6282874 (2014)
[6] H. Balaska and S. Ladaci, “Fractional order output-feedback tube-MRAC design for a class of fractional order transfer functions with unknown parameters”, International Journal of Automation and Control 2023 Vol. 17 Issue 3 Pages 287-305(2023)
[7] Y. Ding, X. G. Ren, X. C. Zhang, X. Liu and X. Wang, “Multi-Phase Focused PID Adaptive Tuning with Reinforcement Learning”, Electronics 2023 Vol. 12 Issue 18(2023)
[8] Y. F. Hanna, A. A. Khater, M. El-Bardini and A. M. El-Nagar, “Real time adaptive PID controller based on quantum neural network for nonlinear systems”, Engineering Applications of Artificial Intelligence 2023 Vol. 126(2023)
[9] L. Garza, A. Vargas and R. Cruz, “MRAC-based Fault Tolerant Control of a SISO Real Process Application”, IEEE Latin America Transactions 2015 Vol. 13 Issue 8 Pages 2545-2550(2015)
[10] R. Dey, S. K. Jain and P. K. Padhy, “Robust closed loop reference MRAC with PI compensator”, Iet Control Theory and Applications 2016 Vol. 10 Issue 18 Pages 2378-2386(2016)
[11] H. Fu, X. Chen, W. Wang and M. Wu, “MRAC for unknown discrete-time nonlinear systems based on supervised neural dynamic programming”, Neurocomputing 2020 Vol. 384 Pages 130-141(2020)
[12] M. Manimozhi and A. A. Rajathi, “Design of MRAC and modified MRAC for DC motor speed control”, International Journal of Nonlinear Analysis and Applications 2021 Vol. 12 Pages 1863-1871(2021)
[13] X. D. Lv, G. M. Zhang, Z. Q. Bai, X. X. Zhou, Z. H. Shi and M. X. Zhu, “Adaptive Neural Network Global Fractional Order Fast Terminal Sliding Mode Model-Free Intelligent PID Control for Hypersonic Vehicle's Ground Thermal Environment”, Aerospace 2023 Vol. 10 Issue 9(2023)
[14] Ardupilot, http://ardupilot.org/
[15] Mission Planner, http://ardupilot.org/planner/