期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2019
卷号：97
期号：1
页码：183-194
出版社：Journal of Theoretical and Applied
摘要：There are several delivery service problems in companies that have multiple stores in one city. These problems occur especially for companies that offer products that these products must be delivered to the customers� location by using their own delivery service. For several companies, they distribute their stock in a single main warehouse and in their stores. In the other side, their delivery service fleet is also distributed in their main warehouse and in every store. This condition triggers inefficiency in stock and the delivery fleet. In this work, we propose the centralized shared delivery service model. As a centralized model, the delivery service is handled by the central management so that coordination in delivery process among vehicles can be more efficient. As a shared system, the vehicle is not dedicated for single store only so that the vehicle can deliver products that come from more than one store in a single trip. In warehouse management, we use single warehouse concept so that all purchased products from all stores will be delivered from the main warehouse. In this work, we propose modified k-means clustering model in managing the delivery process. By using clustering mechanism, each vehicle will deliver products that their destination location is near to each other. In this work, we propose two variants of the k-means clustering model. In the first variant, we combine the k-means clustering method with the round robin method. In the second variant, we combine the k-means clustering method with sequential vehicle creation method. There are research findings after we have done tests. The increasing of the city size makes all observed variables increase. This condition occurs in all models. The increasing of the maximum delivery distance does not affect the total delivery distance but makes the number of vehicles decrease and in the other side makes the delivery distance per vehicle increase. The increasing of the number of stores does not affect the total delivery distance. In the first model, the increasing of the number of stores makes the number of vehicles increase and the delivery distance per vehicle decrease. In the other models, the increasing of the number of stores does not affect the number of vehicles and the delivery distance per vehicle. The increasing of the number of destinations makes all observed variables increase.
关键词：Delivery Service; K-Means Clustering; Round Robin; Single Warehouse Multi Store.
