Muriyatmoko, Dihin (2026) Pemodelan Dan Penyelesaian Permasalahan Penjadwalan Mata Kuliah Dan Rute Kendaraan Penjemputan Dosen Yang Terintegrasi Menggunakan Algoritma Metaheuristik. Doctoral thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 7026221002-Doctoral.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
7026221002-Doctoral.pdf - Accepted Version Restricted to Repository staff only Download (11MB) | Request a copy |
Abstract
Permasalahan optimasi lintas domain yang mengintegrasikan penjadwalan dan rute kendaraan telah banyak diterapkan di sektor logistik, kesehatan, dan manufaktur, namun masih jarang dikaji di sektor pendidikan tinggi dengan karakteristik multi-lokasi. Kesenjangan ini semakin nyata pada universitas yang beroperasi di beberapa lokasi geografis terpisah, di mana jadwal mata kuliah dan rute kendaraan penjemputan dosen saling memengaruhi dan melibatkan sumber daya bersama, sehingga penyelesaian keduanya secara terpisah menghasilkan solusi yang kurang optimal. Universitas Darussalam (UNIDA) Gontor merupakan contoh nyata dari kondisi tersebut dengan lima kampus yang tersebar di Ponorogo, Ngawi, Kediri, Magelang, dan Banyuwangi. Penelitian ini mengembangkan dua model optimasi berbasis Integer Linear Programming (ILP) untuk menyelesaikan permasalahan secara terintegrasi. Model pertama ditujukan untuk program pascasarjana dengan pendekatan integrasi parsial antara penjadwalan mata kuliah dan penentuan rute kendaraan berkapasitas terbatas. Model kedua ditujukan untuk program sarjana dengan pendekatan integrasi simultan yang menggabungkan penjadwalan mata kuliah dan penugasan armada kendaraan heterogen pada rute tetap. Kedua model diintegrasikan dalam arsitektur optimasi dua tingkat, di mana solusi pascasarjana menjadi batasan tambahan bagi model sarjana. Metode penyelesaian diusulkan dua tahap yaitu Greedy Randomized Constructive Heuristic (GRCH) untuk menghasilkan solusi awal yang memenuhi seluruh batasan keras (hard constraints (HC)), dan Tabu Search (TS) untuk meningkatkan kualitas solusi berdasarkan batasan lunak (soft constraints (SC)). Eksperimen dilakukan menggunakan empat dataset yang mencakup dua semester, satu program pascasarjana dan tujuh fakultas sarjana, dengan sepuluh pengujian independen pada setiap dataset. Hasil penelitian menunjukkan bahwa semua solusi memenuhi HC 100%. Pada kasus pascasarjana, semua SC terpenuhi dengan penghematan biaya transportasi sebesar 25,0% dibandingkan pendekatan sekuensial, serta menghasilkan kualitas solusi yang sebanding dengan solver eksak. Pada kasus sarjana, menghasilkan penghematan biaya sebesar 27,8% dibandingkan solver eksak, serta mengungkap keterbatasan fungsi bobot tetap dalam merepresentasikan biaya riil. Selain itu, Algoritma TS unggul dalam efisiensi komputasi, enam kali lebih cepat dari solver eksak. Penelitian ini memberikan tiga kontribusi yaitu formulasi permasalahan integrasi baru pada konteks penjadwalan akademik dan transportasi dosen di universitas multi-lokasi, pengembangan arsitektur optimasi dua tingkat yang menggabungkan integrasi parsial dan simultan, serta bukti empiris penghematan biaya operasional sebagai dasar implementasi pendekatan metaheuristik
========================================================================================================================================
Cross-domain optimization problems integrating course timetabling and vehicle routing have been widely studied in logistics, healthcare, and manufacturing sectors; however, they remain underexplored in higher education, particularly in multi-campus settings. This research gap becomes more evident in universities operating across geographically dispersed locations, where course schedules and lecturer transportation routes are interdependent and share common resources. Solving these problems separately often leads to suboptimal solutions. Universitas Darussalam (UNIDA) Gontor represents a real-world case of such conditions, operating across five campuses located in Ponorogo, Ngawi, Kediri, Magelang, and Banyuwangi. This study develops two optimization models based on Integer Linear Programming (ILP) to address the problem in an integrated manner. The first model is designed for postgraduate programs using a partial integration approach that combines course timetabling with vehicle routing decisions under limited capacity constraints. The second model is developed for undergraduate programs using a simultaneous integration approach that jointly optimizes course timetabling and the assignment of a heterogeneous vehicle fleet to fixed routes. Both models are connected within a two-level optimization architecture, where the postgraduate solution serves as an additional constraint for the undergraduate model. A two-stage solution approach is proposed. The first stage employs a Greedy Randomized Constructive Heuristic (GRCH) to generate an initial feasible solution satisfying all hard constraints (HC). The second stage applies Tabu Search (TS) to iteratively improve solution quality based on soft constraints (SC). Computational experiments are conducted on four real-world datasets covering two academic semesters, involving one postgraduate program and seven undergraduate faculties, with ten independent runs for each dataset. The results show that all solutions satisfy the hard constraints with a 100% feasibility rate. For the postgraduate case, all soft constraints are satisfied, achieving a 25.0% reduction in transportation cost compared to a sequential approach, while maintaining solution quality comparable to an exact solver. For the undergraduate case, the proposed approach achieves a 27.8% cost reduction compared to the exact solver and reveals limitations of fixed-weight cost functions in representing real operational costs. Furthermore, the TS algorithm demonstrates superior computational efficiency, being approximately six times faster than the exact solver. This study contributes by proposing a novel integrated optimization problem in the context of academic scheduling and lecturer transportation in multi campus universities, developing a two-level optimization architecture combining partial and simultaneous integration strategies, and providing empirical evidence of significant operational cost savings to support the implementation of metaheuristic approaches.
| Item Type: | Thesis (Doctoral) |
|---|---|
| Uncontrolled Keywords: | integrasi lintas domain, penjadwalan mata kuliah, rute kendaraan, metaheuristik, universitas multi-lokasi, cross-domain integration, course timetabling, vehicle routing, metaheuristic, multi-location university |
| Subjects: | T Technology > T Technology (General) > T57.6 Operations research--Mathematics. Goal programming |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 55003-(S3) PhD Thesis |
| Depositing User: | Dihin Muriyatmoko |
| Date Deposited: | 26 Jul 2026 04:44 |
| Last Modified: | 26 Jul 2026 04:44 |
| URI: | http://repository.its.ac.id/id/eprint/137307 |
Actions (login required)
 |
View Item |