Metaheuristics for multiobjective optimisation

This work deals with the design, implementation and experimental analysis of metaheuristics for solving multiobjective optimisation problems, with a particular interest on hard and large combinatorial problems from the field of logistics. After focusing on a unified view of multiobjective metaheuristics, we propose new cooperative, adaptive and parallel approaches. The performance of these methods are experimented on a scheduling and a routing problem involving two or three objective functions. We finally discuss how to adapt such metaheuristics during the search process in order to handle uncertainty that may occur from many different sources.