Abstract
The on-line vehicle routing problems (VRP) is an extension of VRP in order to consider real-time requests as well as possible variations travel times in the network. In this research, a solution algorithm for solving on-line VRP is proposed. The solution algorithm is divided into two phases, off-line route planning and on-line route updating. In the off-line phase, a time-dependent VRP formulation is constructed to assign initial routes. In the on-line routing phase, a hybrid heuristic approach with tabu search and genetic algorithm is proposed to handle real-time requests and to improve routes under real-time information. The simulation-assignment model, DynaTAIWAN is applied to evaluate assigning and routing strategies in a traffic network. Numerical experiments are conducted in a Kaohsiung city network.
