Monte Carlo simulation and genetic algorithm for optimising supply chain management in a stochastic environment

Olfa Jellouli (1) , Eric Chatelet (1)

1 LM2S - Laboratoire Modélisation et Sûreté des Systèmes

Open market and e-commerce change the environment of the manufacturing system. In fact, nowadays, vendors are facing a more and more flexible demand. At the same time, we have more accurate tools to study demand evolution and characteristics. This paper reports a methodology to adopt for optimizing the supply chain inventory management (SCIM) taking into account parameters characterizing this uncertain environment. This approach tends to reduce the cost of parameters changing when dealing with a flexible demand. We focus on the management of stochastic parameters characterizing the demand such as stochastic lead time, quantity and rate and other parameters such as delivery time. The objective fixed is to optimize the profit composed of unsatisfied demand, backlog, inventory and production costs. Since an analytical formula of the profit isn't possible, we use Monte Carlo simulation and genetic algorithms. Numerical results are given for two cases.

Domains: Operations Research [cs.RO]; Operator Algebras [math.OA]

Olfa Jellouli, Eric Chatelet. Monte Carlo simulation and genetic algorithm for optimising supply chain management in a stochastic environment. IEEE International Conference on Systems, Man & Cybernetics, Oct 2001, Tucson, United States. pp.1835-1839, ⟨10.1109/ICSMC.2001.973596⟩. ⟨hal-02353049⟩