Optimisasi Strategi Distribusi Barang Antar Pulau Menggunakan Clustering dan Hybrid Adaptive Large Neighborhood Search

Galih, Puji Ridho (2025) Optimisasi Strategi Distribusi Barang Antar Pulau Menggunakan Clustering dan Hybrid Adaptive Large Neighborhood Search. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6032231141-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Distribusi barang antarpulau melalui Tol Laut dari Tanjung Perak, Surabaya, ke Indonesia Timur menghadapi tantangan efisiensi akibat rendahnya muatan balik, yang menyebabkan utilisasi kapasitas kapal rendah. Penelitian ini bertujuan mengoptimalkan distribusi barang dengan merancang Jaringan Pasok Terdesentralisasi menggunakan K-Means Clustering, manajemen inventaris berbasis model EOQ-ROP stokastik, dan Capacitated Vehicle Routing Problem (CVRP) berbasis algoritma Hybrid Adaptive Large Neighbourhood Search (HALNS).
Sebanyak 96 pelabuhan dikelompokkan menjadi 5 klaster menggunakan K-Means Clustering, dipilih berdasarkan Composite Score tertinggi (0,6363; Silhouette Score 0,4261, Davies-Bouldin Index 0,9047, Calinski-Harabasz Index 94,545). Gudang sentral per klaster ditentukan melalui metrik sentralitas jaringan (Closeness Centrality, Connectivity) dan prioritas pelabuhan, menghasilkan Merauke, Makassar, Sorong, Bitung, dan Kupang sebagai hub strategis. Manajemen inventaris berbasis EOQ-ROP stokastik den gan faktor musiman dan simulasi Monte Carlo diterapkan untuk mengelola pesanan secara efisien. Analisis Pareto mengidentifikasi 20% komoditas dengan permintaan tertinggi, mendukung pengembangan industri lokal untuk meningkatkan muatan balik.
CVRP dengan HALNS mengoptimalkan rute pada distribusi primer (Tanjung Perak–gudang sentral) dan sekunder (gudang sentral–pelabuhan tujuan), menghasilkan pengurangan biaya operasional sebesar 22,5% (dari $10.200.510,00 menjadi $7.909.694,50) dan peningkatan utilisasi kapal dari 39,3% menjadi 53,1% dibandingkan sistem Tol Laut 2024. Jaringan Pasok Terdesentralisasi ini meningkatkan efisiensi logistik, menekan biaya, dan mendukung pemerataan ekonomi di Indonesia Timur melalui pengembangan industri lokal.
=======================================================================================================================================
Inter-island goods distribution via the Tol Laut program from Tanjung Perak, Surabaya, to Eastern Indonesia faces efficiency challenges due to low return cargo, resulting in suboptimal vessel capacity utilization. This study optimizes distribution by designing a Decentralized Supply Network using K-Means Clustering, stochastic EOQ-ROP inventory management, and Capacitated Vehicle Routing Problem (CVRP) with the Hybrid Adaptive Large Neighbourhood Search (HALNS) algorithm.
Ninety-six ports are grouped into five clusters using K-Means Clustering, selected for its highest Composite Score of 0.6363 (Silhouette Score 0.4261, Davies-Bouldin Index 0.9047, Calinski-Harabasz Index 94.545). Central warehouses per cluster are determined based on network centrality metrics (Closeness Centrality, Connectivity), identifying Merauke, Makassar, Sorong, Bitung, and Kupang as strategic hubs. Stochastic EOQ-ROP inventory management, incorporating seasonal factors and Monte Carlo simulation, ensures efficient order management under demand and lead time variability. Pareto analysis identifies the top 20% of commodities driving demand, guiding local industry development to enhance return cargo.
CVRP with HALNS optimizes routes for primary (Tanjung Perak to central warehouses) and secondary (central warehouses to destination ports) distribution, achieving a 22.5% reduction in operational costs (from $10,200,510.00 to $7,909,694.50) and increasing vessel utilization from 39.3% to 53.1% compared to the 2024 Tol Laut system. This Decentralized Supply Network enhances logistics efficiency, reduces costs, and promotes economic equity in Eastern Indonesia through local industry growth.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	HALNS, K-Means, Pareto Analysis, Stochastic EOQ
Subjects:	T Technology > T Technology (General) T Technology > T Technology (General) > T57.6 Operations research--Mathematics. Goal programming T Technology > T Technology (General) > T57.84 Heuristic algorithms.
Divisions:	Interdisciplinary School of Management and Technology (SIMT) > 61101-Master of Technology Management (MMT)
Depositing User:	Puji Ridho Galih
Date Deposited:	28 Jul 2025 10:14
Last Modified:	28 Jul 2025 10:14
URI:	http://repository.its.ac.id/id/eprint/122893

Actions (login required)

View Item