Ramadhan, Fikri (2026) Aplikasi Algoritma CAOA-SSR untuk Optimasi Job Shop Scheduling Problem pada Jaringan Pelayaran Liner Kapal Penumpang PT. PELNI (Persero). Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5002221001-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5002221001-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy |
Abstract
Penjadwalan voyage kapal penumpang pada jaringan pelayaran liner perlu mempertimbangkan urutan layanan pelabuhan, kapasitas berth, waktu layar antarpelabuhan, dan kendala pasang surut. Penelitian ini memformulasikan penjadwalan voyage PT. PELNI (Persero) sebagai varian Job Shop Scheduling Problem (JSSP), dengan voyage sebagai job, kunjungan pelabuhan sebagai operasi, serta pelabuhan atau fasilitas layanan sebagai machine; fungsi objektifnya adalah minimisasi total tardiness pada level voyage. Solusi direpresentasikan sebagai vektor prioritas kontinu berbasis random-key, sedangkan decoder deterministik membangun jadwal feasible melalui pemilihan eligible operation, pencarian earliest feasible slot, serta pemeriksaan kapasitas dan tidal feasible window. Algoritma usulan, CAOASSR, merupakan hibridasi algoritma Crocodile Ambush Optimization Algorithm (CAOA) dengan Search Space Reduction (SSR) sebagai search guidance berbasis elite dan knowledge-guided reinitialization. Eksperimen menggunakan dataset Januari yang mencakup 624 operasi, 44 pasangan job_id--voyage, dan 72 machine. Teknik first-come first-served (FCFS) digunakan sebagai baseline operasional deterministik, sedangkan algoritma Grey Wolf Optimizer (GWO), CAOA, dan CAOASSR dievaluasi pada skenario N=5, N=10, dan N=15, masing-masing 30 run dengan seed 1--30. Pada hasil representatif, CAOASSR memperoleh total tardiness 265 jam, lebih rendah daripada FCFS 359 jam, GWO 277 jam, dan CAOA 269 jam. Pada evaluasi multi-run, CAOASSR menghasilkan rata-rata total tardiness terendah dan standar deviasi terkecil pada seluruh skenario; pada N=15, nilainya 265,066667 dengan standar deviasi 0,253708. Uji Friedman pada N=15 menunjukkan perbedaan performa signifikan (p=3,17 x 10^-6), dan uji Wilcoxon dengan koreksi Holm menunjukkan CAOASSR berbeda signifikan terhadap CAOA dan GWO; ukuran efek Vargha--Delaney A12 terhadap GWO sebesar 0,8422 dan berkategori besar. Seluruh 270 run stokastik menghasilkan jadwal feasible.
===============================================================================================================================
Passenger voyage scheduling in a liner shipping network must consider port-service sequences, berth capacity, inter-port sailing times, and tidal constraints. This study formulates the voyage scheduling problem of PT. PELNI (Persero) as a variant of the Job Shop Scheduling Problem (JSSP), with voyages as jobs, port calls as operations, and ports or service facilities as machines; the objective function minimizes total tardiness at the voyage level. Solutions are represented as continuous random-key-based priority vectors, while a deterministic decoder constructs feasible schedules through eligible-operation selection, earliest feasible slot search, and capacity and tidal feasible-window checks. The proposed algorithm, CAOASSR, hybridizes the Crocodile Ambush Optimization Algorithm (CAOA) with Search Space Reduction (SSR) as elite-based search guidance and knowledge-guided reinitialization. The experiments use a January dataset comprising 624 operations, 44 job_id–voyage pairs, and 72 machines. The first-come first-served (FCFS) technique is used as a deterministic operational baseline, while the Grey Wolf Optimizer (GWO), CAOA, and CAOASSR are evaluated under the N = 5, N = 10, and N = 15 scenarios, each with 30 runs using seeds 1–30. In the representative results, CAOASSR obtains a total tardiness of 265 hours, lower than FCFS at 359 hours, GWO at 277 hours, and CAOA at 269 hours. In the multi-run evaluation, CAOASSR achieves the lowest mean total tardiness and the smallest standard deviation across all scenarios; at N = 15, the values are 265.066667 and 0.253708, respectively. The Friedman test at N = 15 indicates a significant performance difference (p = 3.17 × 10^-6), and the Wilcoxon test with Holm correction shows that CAOASSR differs significantly from CAOA and GWO; the Vargha–Delaney A12 effect size against GWO is 0.8422 and is categorized as large. All 270 stochastic runs produce feasible schedules.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | CAOASSR, CAOA, Job Shop Scheduling Problem, penjadwalan voyage, random-key encoding, total tardiness, pasang surut, voyage scheduling, tides |
| Subjects: | H Social Sciences > HD Industries. Land use. Labor > HD69.T54 Time management. Scheduling H Social Sciences > HE Transportation and Communications > HE147.6 Transportation--Planning Q Science > Q Science (General) > Q337.3 Swarm intelligence T Technology > T Technology (General) > T57.6 Operations research--Mathematics. Goal programming T Technology > T Technology (General) > T57.84 Heuristic algorithms. |
| Divisions: | Faculty of Mathematics, Computation, and Data Science > Mathematics > 44201-(S1) Undergraduate Thesis |
| Depositing User: | Fikri Ramadhan |
| Date Deposited: | 29 Jul 2026 02:16 |
| Last Modified: | 29 Jul 2026 02:16 |
| URI: | http://repository.its.ac.id/id/eprint/139024 |
Actions (login required)
 |
View Item |