Manufacturing network configuration by optimal scheduling considering specific machines and alternative process plans

Formation of a networked manufacturing system for particular multi-part product orders is tackled in the presence of alternative process plans for the parts. Each process plan can be transformed into a production schedule in different ways, depending on the machines that are available for each operation in various nodes of the network. Machine suitability is decided on technical criteria using an appropriate database. Optimum assignment of machines to operations is achieved by a Genetic Algorithm (GA), its chromosome representing all operations of the process plans with integer encoding. The GA evaluation function for each network configuration solution uses total order completion time, derived by automatic construction corresponding Gantt charts, subtracting bonuses to promote combined transportation of individual part batches and adding penalties for delays and use of specific highly loaded machines. An example is presented and a methodology for investigating GA performance enhancement is carried out.