Abstract

This research tackles a supplier selection problem composed of different suppliers with limited capacities, a client with deterministic multi-period demands and specific allowed inventory limit in each period for a single product. The objective is to select the most economical set of suppliers in order to meet the client's demand. A novel genetic algorithm and chromosome representation are proposed to find near optimal solutions. The performance of the proposed algorithm is compared with the exact approach using randomly generated data sets. In this supplier selection problem, initially proper population size is determined for three different problem sizes of small, medium and large; as the next step of the experiments, proper numbers of iterations for each problem size are found; finally, different mutation probabilities are tested for different problem sizes and the best mutation probabilities for each problem size are selected based on the calculated error. By the help of the results of the experiments and gathered information, proper population size, number of iterations, and mutation probabilities are recommended for problems with similar size and constraints. Key words: Supply Chain Management, Supplier Selection, Genetic algorithms, Optimization