Solving the Manufacturing Cell Design Problem via Invasive Weed Optimization

Manufacturing plants are commonly organized in cells containing machines that process different parts of a given product. The Manufacturing Cell Design Problem (MCDP) aims at efficiently organizing the machines into cells in order to increase productivity by minimizing the inter-cell moves of parts. In this paper, we present a new approach based on Invasive Weed Optimization (IWO) for solving such a problem. The IWO algorithm is a recent metaheuristic inspired on the colonization behavior of the invasive weeds in agriculture. IWO represents the solutions as weeds that grow and produce seeds to be randomly dispersed over the search area. We additionally incorporate a binary neighbor operator in order to efficiently handle the binary nature of the problem. The experimental results demonstrate the efficiency of the proposed approach which is able to reach several global optimums for a set of 90 well-known MCDP instances. R. Soto (✉) ⋅ B. Crawford ⋅ C. Castillo Pontificia Universidad Católica de Valparaíso, Valparaíso, Chile e-mail: [email protected] B. Crawford e-mail: [email protected] C. Castillo e-mail: [email protected] R. Soto Universidad Autónoma de Chile, Santiago, Chile R. Soto Universidad Cientifica del Sur, Lima, Peru B. Crawford Universidad Central de Chile, Santiago, Chile B. Crawford Universidad San Sebastián, Santiago, Chile F. Paredes Escuela de Ingeniería Industrial, Universidad Diego Portales, Santiago, Chile e-mail: [email protected] © Springer International Publishing Switzerland 2016 R. Silhavy et al. (eds.), Artificial Intelligence Perspectives in Intelligent Systems, Advances in Intelligent Systems and Computing 464, DOI 10.1007/978-3-319-33625-1_11 115