This paper describes a general multi-objective mathematical model for the dial-a-ride problem approximately solved by Simulated Annealing. The model deals with a static problem and includes several distinct cases such as heterogeneous or homogeneous fleet of vehicles, multi or single depot and a multi-objective function that treats transportation costs and customer inconveniences by using penalties. The simulated annealing application is straightforward with three types of neighbors' moves that are randomly selected and equally distributed. The routes are clustered and scheduled in a separate way using specific heuristic methods. Computational results are performed over instances of the literature and the results are compared against current state of the art methods.
Dial-a-ride; simulated annealing; multi-objective model
