Genetic algorithm combined with Taguchi method for optimisation of supply chain configuration considering new product design

In this paper, we propose a methodology to optimally configure a supply chain when considering a new product design. The supply chain configuration is conducted during the product design phase. In fact, several product design alternatives are possible and the aim is to select the best product design optimising the supply chain and satisfying market place as well. In this design problem, specificities of the new product architecture and logistical constraints of supply chain partners are considered at the same time. This product-supply chain design process simultaneity is modelled using an UML sequence diagram. Supply chain design is achieved by levels corresponding to the product's bill of material. A mathematical model is proposed for optimising costs for each level. Genetic algorithms are used to solve the complexity of the model. Since parameters values of genetic algorithms have a significant impact on their efficiency, we have proposed to combine Taguchi experimental design and genetic algorithm to determine the optimal combination of parameters that optimises the objective function. This method can effectively reduce time spent on parameter design using genetic algorithm and increase also its efficiency. The accuracy of the proposed GA-Taguchi method is validated using CPLEX software to evaluate its performance.