In this thesis, we focused on the development of heuristic approaches for solvingvehicle routing problems. We exploited researches conducted on interval graphsand dominance properties of saturated tours to deal more efficiently with selectivevehicle routing problems. An adaptation of a particle swarm optimization algorithmand a memetic algorithm is proposed. The metaheuristics that we developed arebased on effective techniques such as optimal split, genetic crossover operatorsand local searches. We are also interested in classical vehicle problems with timewindows. Various pre-processing methods are introduced to obtain lower boundson the number of vehicles. These methods are based on many approaches usinggraph models, scheduling problems and bin packing problems with conflicts. Wealso showed the effectiveness of the developed methods with an industrial applicationby implementing a portal of mobility services.
