Sistem Pendaratan Kooperatif Quadcopter Pada USV dengan Gangguan Angin Dan Gelombang Menggunakan Nonlinear Model Predictive Control

Wisudawati, Setyo Nuraini (2026) Sistem Pendaratan Kooperatif Quadcopter Pada USV dengan Gangguan Angin Dan Gelombang Menggunakan Nonlinear Model Predictive Control. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6022241080_Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (18MB) | Request a copy

Abstract

Pendaratan quadcopter secara otonom pada Unmanned Surface Vehicle (USV) yang bergerak merupakan permasalahan kompleks karena memerlukan koordinasi presisi antara kedua vehicle selama proses pendaratan. Penelitian ini merancang metode pendaratan berbasis Nonlinear Model Predictive Control (NMPC) kooperatif, dimana quadcopter dan USV saling bertukar informasi posisi melalui komunikasi, dengan dinamika kedua vehicle dimodelkan dalam bentuk nonlinear 6-DOF. Karena kecepatan translasi tidak dapat diukur secara langsung, Extended Kalman Filter (EKF) digunakan untuk mengestimasi state tersebut sekaligus berperan sebagai prediktor horizon NMPC. Gangguan angin terhadap quadcopter dimodelkan menggunakan turbulensi Dryden, sedangkan gangguan gelombang terhadap USV dimodelkan menggunakan spektrum JONSWAP. Pengujian dilakukan melalui simulasi pada MATLAB R2024b dengan lima skenario berdurasi 60 detik yaitu tanpa gangguan lingkungan, dengan gangguan angin, dengan gangguan gelombang, dengan gangguan angin dan gelombang secara bersamaan, serta communication lost. Hasil simulasi skenario tanpa gangguan lingkungan menunjukkan pendaratan berhasil pada t = 24,1 detik dengan galat posisi horizontal kecil. Pada skenario gangguan angin dengan total 20 seed, pendaratan tercapai pada rentang 26,9 – 28,9 detik dengan rata-rata 27,49 detik, dengan error dominan pada sumbu vertikal yaitu RMSE 1,25 – 1,30 m dibanding horizontal yaitu RMSE 0,24 – 0,42 m . Pada skenario gangguan gelombang dengan total 17 seed, dengan waktu pendaratan bervariasi 19,1 – 31,3 detik dengan rata-rata 25,05 detik, akibat sifat stokastik fase gelombang. Pada skenario gabungan angin dan gelombang dengan total 14 seed, sistem tetap berhasil mendarat pada rentang 19,7 – 29,4 detik, dengan rata-rata 24,74 detik, meski galat horizontal meningkat akibat interaksi dua gangguan sekaligus. Pada skenario communication lost, deviasi posisi horizontal sementara hingga 1 – 1,5 meter terjadi selama komunikasi terputus, namun sistem berhasil memulihkan pelacakan dan menyelesaikan pendaratan begitu komunikasi tersambung kembali. Secara keseluruhan, metode NMPC-EKF terbukti mampu menyelesaikan pendaratan pada seluruh seed yang diuji di berbagai kondisi gangguan lingkungan maupun gangguan komunikasi.

============================================================
==========================================================================

Autonomous landing of a quadcopter on a moving Unmanned Surface Vehicle (USV) is a complex problem, as it requires precise coordination between the two vehicles throughout the landing process. This research develops a landing method based on cooperative Nonlinear Model Predictive Control (NMPC), in which the quadcopter and USV exchange position information through communication, with the dynamics of both vehicles modeled as nonlinear 6-DOF systems. Since translational velocity cannot be measured directly, an Extended Kalman Filter (EKF) is employed to estimate these states while also serving as the NMPC horizon predictor. Wind disturbance on the quadcopter is modeled using Dryden turbulence, while wave disturbance on the USV is modeled using the JONSWAP spectrum. Testing was conducted through simulation in MATLAB R2024b across five 60-second scenarios: without environmental disturbance, with wind disturbance, with wave disturbance, with combined wind and wave disturbance, and with communication loss. Simulation results for the scenario without environmental disturbance show that landing was successfully achieved at t = 24.1 s with small horizontal position error. In the wind disturbance scenario, tested across 20 seeds, landing was achieved within a range of 26.9–28.9 s with an average of 27.49 s, with error dominated by the vertical axis RMSE 1.25–1.30 m compared to the horizontal axes RMSE 0.24–0.42 m. In the wave disturbance scenario, tested across 17 seeds, landing time varied between 19.1–31.3 s with an average of 25.05 s, resulting from the stochastic nature of the wave phase. In the combined wind and wave disturbance scenario, tested across 14 seeds, the system still successfully achieved landing within a range of 19.7–29.4 s, with an average of 24.74 s, although horizontal error increased due to the interaction of the two simultaneous disturbances. In the communication loss scenario, a temporary horizontal position deviation of up to 1–1.5 m occurred during the communication outage, but the system successfully recovered tracking and completed the landing once communication was restored. Overall, the cooperative NMPC method based on EKF estimation proved capable of successfully completing landing across all tested seeds under various maritime environmental and communication disturbance conditions.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Dryden, EKF, JONSWAP, NMPC, pendaratan quadcopter, Dryden, EKF, JONSWAP, NMPC, quadcopter landing
Subjects:	T Technology > TJ Mechanical engineering and machinery > TJ217.6 Predictive Control T Technology > TL Motor vehicles. Aeronautics. Astronautics > TL776 .N67 Quadrotor helicopters--Automatic control U Military Science > UG1242 Drone aircraft--Control systems. (unmanned vehicle) V Naval Science > VM Naval architecture. Shipbuilding. Marine engineering > VM365 Remote submersibles. Autonomous vehicles.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User:	Setyo Nuraini Wisudawati
Date Deposited:	01 Aug 2026 05:29
Last Modified:	01 Aug 2026 05:29
URI:	http://repository.its.ac.id/id/eprint/142785

Actions (login required)

View Item