Planning the reverse logistic for recyclable solid waste by a mathematical model and a simulated annealing metaheuristic

Selective collection has grown over the last few years and this is leading to an increase of the reverse logistics costs paid by the city hall. Thus, a better planning of the reverse logistics may reduce costs by minimising the number of trucks needed and the total distance travelled. This paper proposes a mathematical model and a modified simulated annealing metaheuristic based on the vehicle routing problem with multiple trips to plan the reverse logistics for the selective collection. The mathematical model was solved by CPLEX 12.6 and the SA was developed in C. In order to test both of them, they were applied to the selective collection in five Brazilian cities that use a volunteer delivery stations (VDS) located at roads and squares. SA performed the planning of selective collection in a relatively low computational time and achieved a better plan than the ones done manually by the planners.