A Hybrid Tabu Search Algorithm for the Variable Periodic Vehicle Routing Problem

This paper presents a real-life variable periodic vehicle routing problem (VPVRP) which is originated from an auto parts company. The idea of lean production has been applied in the production of the auto parts manufacturer. The VPVRP consists of the assignment of a pickup schedule for each of its supplier's locations and also the establishment of daily routes in order to minimize total transportation cost. Since the problem is a special case of the periodic vehicle routing problem, several initial solution algorithms and a hybrid tabu search (HTS) are developed for solving the problem. There are distinctive features in the HTS, including the intensification strategy, the diversification strategy, and the combination of tabu search with the switch mechanism. The computational results show that the proposed initial algorithms and the HTS outperform the current scheduling method used by the case plant with a significant improvement.