The Formation of Manufacturing Cells Using Genetic Algorithms

This paper proposes a genetic-based algorithm to handle the multi-criteria optimisation problem associated in the formation of cells in Group Technology (GT). GT or Cellular Manufacturing (CM) is a concept where a manufacturing system is decomposed into subsystems or cells. This is done by grouping a variety of parts with similar shape, dimension or process route. This manufacturing concept allows small batch-type production to gain economic advantages similar to those in mass production and still retain the the flexibility of job-shop production. In this report, a genetic-based algorithm is developed to solve the cell formation problem. Genetic Algorithm (GA) is an optimisation technique that imitates the survival-of-the-fittest concept. The advantages of applying the GA approach in this problem include producing more than one acceptable solution and using several objective functions. To overcome the problem of multi-criteria optimisation associated in the formulation of cells, the criteria are prioritised and modelled as multi-objective functions in the algorithm. Consequently, the algorithm is able to find a compromise between goals. Three different objective functions are used: minimising the inter-cell movement, minimising the variation of workload and maximising the similarity of machines within cells.