DEVELOPING MODEL AND ALGORITHM OF SUPPLIER SELECTION FOR PERISHABLE MATERIAL WITH QUANTITY DISCOUNT AND PAYMENT METHOD
This research presents supplier selection and lot sizing problem (MLSSP) for perishable material with quantity discount and payment method. This problem occured when every supplier of perishable material offer different discount policy. If company buy some material in larger quantity, the price per...
Saved in:
Main Author: | |
---|---|
Format: | Theses |
Language: | Indonesia |
Online Access: | https://digilib.itb.ac.id/gdl/view/24498 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Institut Teknologi Bandung |
Language: | Indonesia |
Summary: | This research presents supplier selection and lot sizing problem (MLSSP) for perishable material with quantity discount and payment method. This problem occured when every supplier of perishable material offer different discount policy. If company buy some material in larger quantity, the price per unit will be cheaper. However, holding cost and perishable cost will be increase. Meanwhile, there is interest charge if total price higher than budget. This kind of problem is faced by PT. Dirgantara Indonesia (PT.DI). In 2016, PT. DI bought non metal material for Airbus Spirit 4% higher than required. As a result, holding cost and interest charge increased to 0,02% and 0,06% of total inventory cost. Therefore, this research will develop MLSSP model to minimize total iventory cost. <br />
<br />
<br />
The inverse function cause model is non linier. Hence, enumeration only satisfy small instance problem. Besides model development, this research will develop algorithm based on simulated anneling (SA). Initial solution is randomly generated with condition only one supplier is choosen to supply every material per period. Initial solution quality then improved by using 3 (three) operators, quantity modified, quantity move, and supplier exchange. Through a series of computational experiments, in average, the algorithm’s performance achieve 93,53% of optimal method with 1,21% computational time. Comparison between proposed method’s best solution dan referenced method’s best solution also indicate proposed method’s fitness function is 2,53% better than referenced method. Meanwhile, proposed method’s computational time is 3,32% faster than referenced method. |
---|