A Heuristic Procedure to Solve an Evolutionary Vehicle Routing Problem with Time Windows

The paper deals with a hybrid clustering-genetic algorithm that is applied to a rolling-horizon procedure for solving the dynamic vehicle routing problem with time windows. The paper mainly focuses on the advantages of introducing a clustering algorithm jointly with the genetic algorithm. Discussion is supported by numerical tests carried out by simulating the implementation of the procedure to the urban logistic centre of Padova, Italy. Results highlight that introducing clustering criteria to guide a genetic algorithm in the operations of generating the initial population, mutation and cross-over can improve the solution as 20%.