Kontrol Trajectory Tracking Hexacopter Menggunakan Adaptive Sliding Mode Control Dan Nonlinear Model Predictive Control

Sawitri, Rany (2026) Kontrol Trajectory Tracking Hexacopter Menggunakan Adaptive Sliding Mode Control Dan Nonlinear Model Predictive Control. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6022212012-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Hexacopter banyak digunakan untuk pemetaan, inspeksi, dan pengiriman, namun kinerja trajectory tracking sering menurun akibat gangguan angin dan beban dinamis. Penelitian ini mengusulkan arsitektur kontrol hierarkis yang mengintegrasikan Nonlinear Model Predictive Control (NMPC) sebagai outer-loop (20 Hz) dan Adaptive Sliding Mode Control (ASMC) sebagai inner-loop (250 Hz). Estimasi gangguan dilakukan dengan menggabungkan Disturbance Observer (DOB) untuk respons cepat dan Radial Basis Function Neural Network (RBF-NN) untuk pembelajaran adaptif pada gangguan yang tidak termodelkan. Hasil estimasi tersebut digunakan sebagai kompensasi feedforward. Evaluasi dilakukan melalui simulasi dengan trajektori lingkaran 3D, yang menunjukkan penurunan RMSE posisi sebesar 37.5%–70.4%, dengan peningkatan tertinggi pada sumbu Z sebesar 82.4%. ASMC meningkatkan kinerja 0.6%–5.2% dibanding SMC konvensional. DOB menunjukkan estimasi lebih akurat (RMSE 0.12 m/s²) dengan response time lebih cepat (~30 ms), sedangkan RBF-NN (RMSE 0.15 m/s², response time ~80 ms) berperan sebagai smoothing filter dan mekanisme adaptif terhadap perubahan karakteristik gangguan
=============================================================================================================================
Hexacopters are widely used for mapping, inspection, and delivery; however, trajectory tracking performance often degrades due to wind disturbances and dynamic payloads. This study proposes a hierarchical control architecture integrating Nonlinear Model Predictive Control (NMPC) as the outer-loop (20 Hz) and Adaptive Sliding Mode Control (ASMC) as the inner-loop (250 Hz). Disturbance estimation is performed by combining Disturbance Observer (DOB) for fast response and Radial Basis Function Neural Network (RBF-NN) for adaptive learning of unmodeled disturbances. The estimation results are used as feedforward compensation. Evaluation is conducted through simulation using a 3D circular trajectory, showing a position RMSE reduction of 37.5%–70.4%, with the highest improvement of 82.4% on the Z-axis. ASMC improves performance by 0.6%–5.2% compared to conventional SMC. DOB demonstrates more accurate estimation (RMSE 0.12 m/s²) with faster response time (~30 ms), while RBF-NN (RMSE 0.15 m/s², response time ~80 ms) serves as a smoothing filter and adaptive mechanism for varying disturbance characteristics.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Hexacopter, NMPC, ASMC, Disturbance Observer, RBF Neural Network, Trajectory Tracking
Subjects:	T Technology > TL Motor vehicles. Aeronautics. Astronautics
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User:	Rany Sawitri
Date Deposited:	04 Feb 2026 06:18
Last Modified:	04 Feb 2026 06:18
URI:	http://repository.its.ac.id/id/eprint/131966

Actions (login required)

View Item