Implementasi ALOS Sebagai Sistem Pemanduan Dan AVO Untuk Penghindaran Rintangan Dinamis Berbasis Perencanaan Lintasan Global Menggunakan Weighted A* Pada USV

Haifa, Alia Nisa (2026) Implementasi ALOS Sebagai Sistem Pemanduan Dan AVO Untuk Penghindaran Rintangan Dinamis Berbasis Perencanaan Lintasan Global Menggunakan Weighted A* Pada USV. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5022221062-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Unmanned Surface Vehicle (USV) memerlukan sistem navigasi yang mampu merencanakan lintasan, mengikuti lintasan, dan menghindari rintangan secara aman agar dapat beroperasi secara otonom pada lingkungan perairan yang dinamis. Pada Tugas Akhir ini dikembangkan sistem navigasi terintegrasi menggunakan Weighted A* sebagai global path planner, G² Continuous Cubic Bézier Spiral (G²-CBS) sebagai metode path smoothing, Adaptive Line-of-Sight (ALOS) sebagai metode path following, dan Adaptive Velocity Obstacle (AVO) sebagai algoritma local obstacle avoidance. Sistem dibangun berdasarkan model dinamika USV 4-DOF hasil identifikasi menggunakan metode Nonlinear Least Squares (NLS), kemudian divalidasi melalui simulasi MATLAB dan diimplementasikan menggunakan Python berbasis Robot Operating System (ROS) pada platform USV LSS-01. Hasil pengujian menunjukkan bahwa algoritma Weighted A* dengan ε = 1,5 memberikan kompromi terbaik antara efisiensi komputasi dan kualitas lintasan, dengan waktu komputasi 0,036 s, expanded node sebanyak 99, dan panjang lintasan 55,181 m. Proses path smoothing menghasilkan lintasan yang lebih kontinu dengan memperpendek lintasan menjadi 51,755 m serta menurunkan mean curvature sekitar 33,3%. Pada pengujian path following, metode ALOS menghasilkan Mean Cross-Track Error (CTE) sebesar 0,0165 m, RMS CTE sebesar 0,0297 m, dan Maximum CTE sebesar 0,1021 m. Pada pengujian local obstacle avoidance, baik metode Velocity Obstacle (VO) maupun AVO berhasil menghindari rintangan tanpa tabrakan. Metode VO menghasilkan akurasi lintasan yang lebih baik dengan Maximum CTE sebesar 0,4468 m, sedangkan AVO menghasilkan manuver yang lebih halus dengan Maximum Roll sebesar 3,6989°, lebih rendah dibandingkan VO sebesar 6,5549°. Hasil implementasi lapangan menunjukkan bahwa USV LSS-01 berhasil mengikuti seluruh waypoint, menghindari rintangan statis tanpa tabrakan, serta menyelesaikan lintasan sepanjang 47,61 m dalam waktu 41,4 detik. Hasil tersebut menunjukkan bahwa sistem navigasi yang dikembangkan memiliki kinerja yang efisien, akurat, stabil, dan andal.
======================================================================================================================================
An Unmanned Surface Vehicle (USV) requires a navigation system capable of planning paths, following trajectories, and avoiding obstacles safely to operate autonomously in dynamic aquatic environments. This final project develops an integrated navigation system consisting of Weighted A* as the global path planner, G² Continuous Cubic Bézier Spiral (G²-CBS) for path smoothing, Adaptive Line-of-Sight (ALOS) for path following, and Adaptive Velocity Obstacle (AVO) for local obstacle avoidance. The system was developed based on a 4-DOF USV dynamic model identified using the Nonlinear Least Squares (NLS) method. The proposed system was validated through MATLAB simulations and implemented using Python on the Robot Operating System (ROS) platform for the LSS-01 USV. The results show that Weighted A* with ε = 1.5 provides the best trade-off between computational efficiency and path quality, achieving a computation time of 0.036 s, 99 expanded nodes, and a path length of 55.181 m. The G²-CBS path smoothing method produces a smoother trajectory by reducing the path length to 51.755 m and decreasing the mean curvature by approximately 33.3%. In the path following evaluation, the ALOS method achieved a Mean Cross-Track Error (CTE) of 0.0165 m, an RMS CTE of 0.0297 m, and a Maximum CTE of 0.1021 m. In the local obstacle avoidance evaluation, both the Velocity Obstacle (VO) and AVO methods successfully avoided obstacles without collisions. The VO method achieved better path-tracking accuracy with a Maximum CTE of 0.4468 m, while the AVO method produced smoother maneuvers with a Maximum Roll of 3.6989°, lower than the 6.5549° obtained by the VO method. Field implementation results demonstrate that the LSS-01 USV successfully followed all waypoints, avoided static obstacles without collisions, and completed a 47.61 m trajectory in 41.4 s. These results indicate that the developed navigation system provides efficient, accurate, stable, and reliable navigation performance for USV applications.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Unmanned Surface Vehicle, Weighted A*, Adaptive Velocity Obstacle, perencanaan lintasan, penghindaran rintangan, path planning, obstacle avoidance
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK3070 Automatic control T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105.546 Computer algorithms
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis
Depositing User:	Alia Nisa Haifa
Date Deposited:	27 Jul 2026 06:46
Last Modified:	27 Jul 2026 06:46
URI:	http://repository.its.ac.id/id/eprint/138167

Actions (login required)

View Item