Scalability in manufacturing systems: a hybridized GA approach

Abstract

As one of the key characteristics in manufacturing systems, scalability plays an increasingly important role that is driven by the rapid change of market demand. It provides the ability to rapidly reconfigure production capacity in a cost-effective manner under different situations. Our industrial partners face scalability problems involving multi-unit and multi-product manufacturing systems. In this paper, a hybridized genetic algorithm (GA) approach is presented to solve these kinds of problems. A mathematical model is defined by considering technological and capacity as well as industrial constraints. Starting from the original process plan and configuration of the manufacturing system, a set of practical principles are built to reduce the time associated with finding a feasible solution. An improved GA is proposed to search in the global solution space; the method is hybridized with a heuristic approach to locally improve the solution between generations. A balancing objective function is defined and used to rank the solutions. Experiments are set to determine the most adequate parameters of the algorithm. An industrial case study demonstrates the validity of the proposed approach.