Multi product reverse logistics network design problem: Mathematical model and simulated annealing based approach

Reverse logistics, which may be defined as picking up end-of-life products from customers and recovering value using appropriate methods, is taking attention both in the literature and in application due to the cost advantages it serves for the companies and also the legislative requirements. In this thesis, a reverse logistics network design problem for multiple products is considered for multiple recovery options. The recovery methods considered are remanufacturing, recycling and elimination of hazardous waste. The main objective of the problem is minimizing total system cost. Furthermore, deciding which potential collection center should be opened at which dimension is another objective of the system. The problem defined in this study distinguishes from other works in the literature because of the fact that it considers recycling, remanufacturing and treatment of hazardous material together, multiple disassembly methods are defined, multiple recyclable material types are defined and multiple parts appropriate for remanufacturing are defined. In order to solve the problem defined in this thesis, two approaches, a mixed integer mathematical model and a simulated annealing based approach are presented. In this context, in order to analyze the efficiency of the proposed solution methods an experimental study which consists of 1350 problems for different network types and for different values of problem parameters is carried out. The results are evaluated and it is found out that the proposed solution methods are effective for the problem defined in this thesis.