Integration of production scheduling and delivery in two echelon supply chain

摘要

For a production scheduling and delivery problem, there are multi-products processed by a distribution centre and then delivered to retailers. The resolution of the problem is to determine the production sequence, retailers' needs to heterogeneous fleet of vehicles and the visiting sequence of each vehicle within time windows. In this article, the problem is considered with minimising the total cost which includes delivery cost, fixed vehicle cost, and penalty cost, subjected to satisfy all demands of each retailer. The adaptive genetic algorithms (AGAs) are designed and tested in variety of production and delivery scenarios. The computational experiments show that backward-AGA (B-AGA) has distinguished performance between two AGAs. However, F-AGA has better solving efficiency when N > 50 and the solving performance is fairly similar subjected to the geographical group R and N > 50. In addition, the experiments indicate that combined with more vehicle types in delivery stage is not only reduced the total cost but also fixed vehicle cost and variable routing cost. Some management recommendations for different map features are provided as well.