A tabu-search heuristic for the vehicle routing problem with self-imposed time windows

In this thesis we analyze the vehicle routing problem with self-imposed time windows and the proposed solution procedure by Jabali et al. (2015). We analyze their results and discuss the differences between their results and the ones we found. In addition to this, we evaluate the effect of different initial solutions, showing that incorporation of shift length in the initial solution improves the results. Furthermore, we implement a random-sized tabu list, which proves to perform almost equally well as a tabu list with a fixed size. Lastly, we extend on the original problem by including time preferences and argue that this increases customer satisfaction.