Perancangan Sistem Kendali Dynamic Obstacle Avoidance Pada Kapal Kargo Umum Berbasis ANN-PSO Dengan Memperhatikan Gangguan

Sari, Melinda 'Jhesika (2025) Perancangan Sistem Kendali Dynamic Obstacle Avoidance Pada Kapal Kargo Umum Berbasis ANN-PSO Dengan Memperhatikan Gangguan. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5009211093-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Selat Gelasa merupakan selat yang memisahkan Pulau Bangka dengan Pulau Belitung. Selat ini memiliki jalur pelayaran dengan tingkat kepadatan yang tinggi, karena berperan sebagai jalur penghubung antara dua perairan yang padat yaitu Laut Natuna Utara (sebagai jalur pelayaran penting) dengan Laut Jawa. Kepadatan lalu lintas pelayaran di Selat Gelasa meningkatkan resiko terjadinya tabrakan antar kapal di wilayah tersebut. Tabrakan antar kapal juga bisa disebabkan oleh human error dan gangguan lingkungan, yang juga berpotensi memperbesar resiko tabrakan antar kapal. Penelitian tugas akhir ini bertujuan untuk merancang dan menganalisis sistem kendali dynamic obstacle avoidance pada kapal kargo umum Mellinda yang berlayar di Selat Gelasa dengan memperhatikan kriteria yang ditetapkan oleh COLREGs. Sistem kendali dirancang menggunakan metode Artificial Neural network (ANN) yang dioptimasi dengan Particle Swarm Optimization (PSO). Simulasi dilakukan menggunakan perangkat lunak MATLAB dan Simulink. Parameter input dari sistem kendali ini berupa TCPA, error kecepatan, error yaw, dan yaw rate. Simulasi sistem dilakukan di beberapa skenario, yaitu ketika kapal berlayar dalam keadaan normal tanpa adanya kapal asing dan gangguan arus laut, ketika terdapat kapal asing yang berpotensi menyebabkan tabrakan tanpa adanya gangguan arus laut, dan dengan adanya gangguan arus laut. Hasil simulasi pertama menunjukkan bahwa dalam keadaan normal, kapal dengan kendali ANN-PSO mampu mengikuti jalur lintasan ideal. Pada simulasi kedua dengan penambahan kapal asing ke dalam simulasi, sistem kendali mampu menghasilkan manuver penghindaran tabrakan ketika kapal berada dalam kondisi beresiko tabrakan dengan kapal asing. Hasil pada simulasi ketiga dengan penambahan gangguan arus laut menunjukkan bahwa kapal dengan kendali ANN-PSO masih bisa melakukan penghindaran tabrakan dengan kapal asing meskipun telah ditambahkan gangguan arus laut.
=======================================================================================================================================
The Gelasa Strait separates Bangka Island from Belitung Island. This strait has a shipping lane with a high level of density, as it serves as a connecting route between two densely trafficked waters: the North Natuna Sea (an important shipping route) and the Java Sea. The high density of shipping traffic in the Gelasa Strait increases the risk of collisions between ships in the area. Collisions between ships can also be caused by human error and environmental disturbances, which also have the potential to increase the risk of collisions between ships. This final project aims to design and analyze a dynamic obstacle avoidance control system on the Mellinda general cargo ship sailing in the Gelasa Strait, taking into account the criteria set by COLREGs. The control system is designed using the Artificial Neural Networks (ANN) method, which is optimized with Particle Swarm Optimization (PSO). The simulation is performed using MATLAB and Simulink software. The input parameters of this control system are TCPA, speed error, yaw error, and yaw rate. The system simulation was conducted in several scenarios: when the ship was sailing under normal conditions without foreign ships and ocean current disturbances, when there were foreign ships that could potentially cause collisions without ocean current disturbances, and with ocean current disturances. The first simulation results show that under normal conditions, ships with ANN-PSO control are able to follow the ideal trajectory. In the seconnd simulation, with the addition of foreign ships to the simulation, the control system was able to generate collision avoidance maneuvers when the ship wa at risk of colliding with foreign ships. The results of the third simulation, with the addition of ocean current diaturbances, show that ships with ANN-PSO control are still able to avoid collisions with foreign ships even when the ocean current disturbances are added.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Dynamic Obstacle Avoidance, ANN-PSO, COLREGs, IMO, TCPA Dynamic Obstacle Avoidance, ANN-PSO, COLREGs, IMO, TCPA
Subjects:	Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science)
Divisions:	Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis
Depositing User:	Melinda 'jhesika Sari
Date Deposited:	05 Aug 2025 06:58
Last Modified:	05 Aug 2025 06:58
URI:	http://repository.its.ac.id/id/eprint/127341

Actions (login required)

View Item