Laboratoire de Génie Informatique et d’Automatique de l’Artois

This thesis deals with the study of an optimization problem in the field of home healthcare. The problem concerns the problem of periodic delivery of consumables to hemodialysis patients at home. The goal is to present an optimization approach that allows decision-makers to estimate the overall cost of a tour with flexible visit frequencies that respect patient storage capacity and visit preference. First, we developed a mathematical model taking into account the specific needs of hemodialysis patients, such as the frequency of deliveries. Then, we developed a variable neighborhood search approach to solve this NP-hard problem, proposing different neighborhood structures and efficient solution schemes. The overall evaluation of the proposed approach on instances from the literature shows the effectiveness of the proposed approach. Thus, the application of the method on a pseudo-real case has shown its
advantages in finding the optimal frequency of visits that respects available patient capacity, time windows and minimizes total costs.