| Summary: | 碩士 === 國立臺灣科技大學 === 工業管理系 === 105 === Facility location and vehicle routing problems are two critical decisions in designing a supply chain network. Although they are often solved separately because of the differences in their planning horizon, these two problems have been proven to be interdependent. Therefore, the location routing problem (LRP) is proposed to solve these problems simultaneously. The problem extend from a practical perspective when the companies do not have their own vehicles to deliver their products. Considering in some real-life situations, arrival time at a customer will affect the customer’s satisfaction level or sales to the customer. Each customer may have a specific service time window during which s/he can be serviced, and a service vehicle must arrive within this time window to start its service. Therefore, this study combines the open location routing problem (OLRP) and time windows constraint into the open location routing problem with time windows (OLRPTW). This study proposes a branch-and-price algorithm to solve OLRPTW. This problem is solved by the simplex algorithm in the master problem. The elementary shortest path problem with resource constraints (ESPPRC) is used in this pricing problem which has the goal to generate feasible routes with negative reduced costs to be added to the restricted master problem (RMP). The OLRPTW instances are derived from three LRP benchmark data sets. The computational result indicates that the branch-and-price algorithm is comparable for small-medium instances. For large instances, it can solve instance with 200 customers and 10 depots to the optimality.
|
|---|
