Simulation and optimisation of a specific flexible manufacturing system.

As current market competition evolves, most companies intend to increase their options for product customisation and accelerate their product upgrading. Correspondingly, manufacturers have to face the increasing size of product family, shortened product life cycle or rapid product/process change. Therefore, Flexible Manufacturing Systems (FMS) have been introduced that uses advanced machines and efficient transport systems to produce multiple products at the same time. However, an FMS can be complicated to manage because of the increased variability in products and processes. The research aims to develop manufacturing simulation and optimisation techniques for a FMS. This research will integrate Discrete Event Simulation (DES) and multi-objective optimisation approach to address the complexity and flexibility within an agile manufacturing environment. Due to the complexity of FMS, most current FMS optimisation research has engaged with FMS production problems separately without considering other inter-related problems in the same system such as dealing with operation sequence problem without considering Level of Flexibility (LoF), thus it is hard for the solution to provide a prospective impact for the whole system. There are very few real-world FMS implementations that are available to literatures, making it difficult to build and verify the models within a complete ecosystem. Consequently, most of the models in the research are oversimplified. Therefore, this research aims to develop a method to optimise FMS production considering the overall system, by having access to an FMS industrial implementation. This research contributes to knowledge in four main areas, namely, (1) the interactions of FMS production problems have been investigated, (2) a framework has been developed to integrate the simulation and optimisation for FMS to enable optimisation algorithms working with DES models effectively, (3) a comprehensive FMS simulation model has been built and validated on the industrial shop floor and (4) multi-objective optimisation has been applied to the FMS scheduling problem, considering interactions with other problems. Based on the results and limitations of this research, real-time simulation, mock-up FMS and improve computational efficiency are suggested for future work.