Implementasi Algoritma Hybrid Cross Entropy–Genetic Algorithm Untuk Menyelesaikan Single Stage Capacitated Warehouse Location Problem (Studi Kasus: PT Petrokimia Gresik)

Lahdji, Fatmah Munif (2016) Implementasi Algoritma Hybrid Cross Entropy–Genetic Algorithm Untuk Menyelesaikan Single Stage Capacitated Warehouse Location Problem (Studi Kasus: PT Petrokimia Gresik). Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[img]
Preview
Text
2512100023-Undergraduate-Theses.pdf

Download (2MB) | Preview

Abstract

Single stage capacitated warehouse location problem (SSCWLP) merupakan permasalahan alokasi distribusi dimana produk akan dikirimkan dari plant menuju warehouse untuk selanjutnya dikirimkan dari warehouse menuju pelanggan. Permasalahan SSCWLP ini merupakan permasalahan NP-hard dimana penyelesaiannya menggunakan metode eksak membutuhkan waktu yang lama sehingga sering didekati dengan metode metaheuristik. Penelitian ini berfokus pada penentuan alokasi distribusi pupuk PT. Petrokimia Gresik di Pulau Sumatera. Penentuan alokasi distribusi pupuk dilakukan menggunakan algoritma hybrid cross entropy – genetic algorithm. Algoritma ini menggabungkan antara mekanisme pembangkitan sampel pada cross entropy dengan mekanisme mutasi pada genetic algorithm sebagai upaya untuk mempercepat waktu komputasi algoritma cross entropy original. Hasil komputasi menggunakan algoritma hybrid cross entropy genetic algorithm menunjukkan penghematan biaya distribusi sebesar Rp 737.780.842,00 dengan membuka empat gudang lini 3, antara lain GP Solok, GP Bukittinggi, GP Merangin, dan GP Kotabumi. Berdasarkan hasil eksperimen yang dilakukan, diketahui bahwa secara umum algoritma hybrid cross entropy – genetic algorithm mampu menghasilkan alternatif solusi yang lebih mendekati optimum dibandingkan algoritma metaheuristik lain seperti algoritma simulated annealing. Selain itu, algoritma hybrid cross entropy – genetic algorithm juga menghasilkan solusi yang lebih mendekati optimum dengan waktu komputasi yang lebih cepat dibangingkan algoritma cross entropy original, dan genetic algorithm tanpa crossover. ================================================================================= Single stage capacitated warehouse location problem (SSCWLP) is an allocation problem in which products will be distributed from the plant to the warehouse for further distributed from the warehouse to the customer. SSCWLP is NP-hard problem, where finding the solution using exact method requires a longer computational time, thus this problem often solved with metaheuristics approach. This study focuses on determining allocation of fertilizer distribution in PT Petrokimia Gresik on Sumatera Island. This allocation problem is computed using a hybrid cross entropy – genetic algorithm method. This algorithm combines the mechanisms of generating samples on cross entropy with mutations mechanism on genetic algorithm in order to speed up the computational time of original cross entropy algorithm. The computational result shows distribution cost savings up to Rp 737.780.842,00 by opening only four gudang lini 3, which are GP Solok, GP Bukittinggi, GP Merangin, and GP Kotabumi.Based on the experimental results, it is known that in general, hybrid cross entropy - genetic algorithm is capable of generating an alternative solution which is closer to the optimum compared to other metaheuristic algorithms such as simulated annealing algorithm. In addition, the hybrid cross entropy algorithm - genetic algorithm also generates solution that is both closer to the optimum solution and has faster computational time compared to original cross entropy algorithms and genetic algorithm without crossover.

Item Type: Thesis (Undergraduate)
Additional Information: RSI 658.5 Lah i
Uncontrolled Keywords: Cross Entropy-Genetic Algorithm; Metaheuristik; Single Stage Capacitated Warehouse Location Problem
Subjects: T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105.546 Computer algorithms
Divisions: Faculty of Industrial Technology > Industrial Engineering > (S1) Undergraduate Theses
Depositing User: Mrs Anis Wulandari
Date Deposited: 15 Jun 2017 04:32
Last Modified: 27 Dec 2018 02:21
URI: http://repository.its.ac.id/id/eprint/41690

Actions (login required)

View Item View Item