Optimasi Stowage Planning Menggunakan Whale Optimization Algorithm Dan Rule-Based Expert System

Wicaksono, Resky Andi (2025) Optimasi Stowage Planning Menggunakan Whale Optimization Algorithm Dan Rule-Based Expert System. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5002211011-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (7MB) | Request a copy

Abstract

Perencanaan penempatan kontainer kapal (Stowage Planning) adalah masalah optimasi kompleks yang melibatkan berbagai batasan, seperti aturan Dangerous Goods, skenario multi-port, massa, dan stabilitas kapal. Penelitian ini mengusulkan model hibrida Whale Optimization Algorithm (WOA) dan Rule-based Expert System (RBES) untuk mencari solusi optimal. WOA dipilih karena kemampuannya menghindari solusi lokal optimal, sementara RBES mengevaluasi tata letak kontainer menggunakan aturan if-then, mencakup kepatuhan regulasi, aksesibilitas multi-port, dan stabilitas kapal seperti LCG (Longitudinal Center of Gravity), TCG (Transversal Center of Gravity), dan VCG (Vertical Center of Gravity). Pelanggaran dihitung sebagai penalti dalam nilai fitness untuk memandu optimasi. Pengujian menggunakan 1.500 kontainer pada kapal 442 TEU (Twenty-foot Equivalent Unit) menunjukkan bahwa model ini berhasil memenuhi sebagian besar batasan, dengan stabilitas horizontal (LCG dan TCG mendekati nol) tercapai, meskipun stabilitas vertikal (VCG) masih fluktuatif. Optimasi menunjukkan konvergensi cepat dengan peningkatan fitness awal dan penurunan keragaman solusi. Pengaturan parameter seperti iterasi dan swap rate memengaruhi hasil akhir. Pendekatan ini terbukti efektif menangani kompleksitas dunia nyata, termasuk Dangerous Goods dan stabilitas kapal, yang belum banyak dieksplorasi sebelumnya.

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

Container stowage planning (Stowage Planning) is a complex optimization problem involving various constraints, such as the Dangerous Goods rule, multi-port scenario, mass, and ship stability. This study proposes a hybrid model of Whale Optimization Algorithm (WOA) and Rule-based Expert System (RBES) to find optimal solutions. WOA was chosen for its ability to avoid locally optimal solutions, while RBES evaluates container layout using if-then rules, encompassing regulatory compliance, multi-port accessibility, and ship
stability constraints such as LCG (Longitudinal Center of Gravity), TCG (Transversal Center of Gravity), and VCG (Vertical Center of Gravity). Violations are calculated as penalties in the fitness value to guide optimization. Testing using 1,500 containers on a 442 TEU (Twenty-foot Equivalent Unit) vessel show that the model successfully meets most constraints, with horizontal stability (LCG
and TCG close to zero) achieved, although vertical stability (VCG) remains volatile. Optimization shows rapid convergence with an increase in initial fitness and a decrease in
solution diversity. Parameter settings such as iterations and swap rates influence the final results. This approach proves effective in handling real-world complexities, including Dangerous Goods and vessel stability, which have not been widely explored before.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Stowage Planning, Whale Optimization Algorithm, Rule-based Expert System, Dangerous Goods, Multi-port, Keseimbangan Kapal ===================================================================================================================================================== Stowage Planning, Whale Optimization Algorithm, Rule-based Expert System, Dangerous Goods, Multi-port, Vessel Stability
Subjects:	Q Science > QA Mathematics > QA336 Artificial Intelligence
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Resky Andi Wicaksono
Date Deposited:	04 Aug 2025 04:28
Last Modified:	04 Aug 2025 04:28
URI:	http://repository.its.ac.id/id/eprint/126775

Actions (login required)

View Item