A multiple objective logistics network model for product recovery and an implementation

Especially within the last decade, Supply Chain Management (SCM), Logistics Management (LM) and Logiştics Network Design (LND) for Product Recovery (PR), have received an increasing attention. LND for PR, is a strategical level planning topic, and related closely with the performance of the system and profitability and sustainability of the business.It has been found a gap in the literature regarding a real application based on LND for PR with multiple objective in uncertain environment. The originality and main contribution and objective of this thesis are the mentioned gap and to realize a value added study on LND for PR which will turn into economical utility and facilatate the human life. A system for LND for Waste Electrical and Electronic Equipment (WEEE), which does not exist yet in Turkey, has been suggested, and its mathematical formulation has been built as deterministic/fuzzy, linear, capacitated, multi-product, multi-objective, mixed integer, location-allocation/LND model. All the districs of Ankara province are taken as customer zones and candidate sites for locating of the collection and treatment facilities. Determining the optimum location, capacity levels and numbers of the facilities to be opened, and allocation of flow levels between/among the customers-collection facilities, collection-treatment and treatment-refining/disposal facilities are considered as decision variables. The Model developed for strategical planning has been solved with a real application using GAMS 22.1 commercial software and CPLEX 8.1 solver. In order to test sensitivity and compare solution results of the Models, over 700 solutions with different parameter values have been done using multiple objective approaches in deterministic and fuzzy environment, and strategic suggestions for efficient and flexible decision making have been presented. For searching the impact and relationship of the variables in the model, independent; one way ANOVA, two way MANOVA and Regression Analysis were implemented using SPSS 17.0 for Windows. Finally, potential and/or new research opportunities are emphasized.