A food restribution model for "la bourse aux dons"

Abstract

This master thesis addresses “La bourse aux dons” problematic about determining the minimum number of vehicles that it would need to acquire in order to satisfy the entire daily “potential donations”. Additionally “La bourse aux dons” wants to measures the environmental and social impact of the delivered surplus food (i.e., potential donations). In the view of the above, it was necessary to determine the transport model that reduces the total cost (fleet and transport cost) satisfying the entire daily potential donations. Therefore, a review on Vehicle Routing Problems (VRPs) and Pickup and Delivery Problems (PDPs) was carried out. The VRPPD was the model that suits best the problem. Then, a review of the VRP’s solution methods in the light of the requirements of “la bourse aux dons” was carried out. Hence, a Grouping genetic algorithm (GGA) was adapted to solve “La bourse aux dons” VRPPD. Furthermore a methodology to find the best GGA’s parameters was performed.

To determine the minimum vehicle fleet size that minimizes the total cost satisfying the entire daily potential donations, a methodology was established based on the solution of multiple random generated samples of the considered sets (20, 30, 50, 100 and 200 potential donations) by using the proposed GGA.

To determine the environmental impact of delivering a considered set, some the water and ecological footprint conversion factors available in the literature were used.

To determine the social impact of delivering a considered set, the number of daily-recommended portions using the considered sets, based on nutritional studies for Belgium.

Based on the discussions and some assumptions, the expected values of the minimum fleet size, economical, environmental and social impact of satisfying the entire daily potential donations were determined.