Heuristic approaches for waste containers location problem and waste collection routes optimisation in an urban area

Abstract

A generalisation of the traditional waste collection vehicle routing problem with time windows (WCVRPTW) is studied in this paper. In case of the number of waste containers is limited, some collection sites requires to be visited for multiple times per day instead of one. Therefore a waste containers management study is required to economise the collection cost. We propose an approach that combines the two problems: containers location (WCLP) and the vehicle routes optimisation. Two algorithms to solve this problem are used. The first is memetic algorithm (MA) to solve the WCLP in order to get the best containers distribution. The second is an iterated local search (ILS) algorithm to solve WCVRPTW with homogenous vehicles and taking into consideration WCLP as constraint. These algorithms are evaluated on classical vehicle routing problem (VRP) instances. The experiment results prove the impact of this combining on collection routes cost in the cited case.