Intelligent flexible manufacturing system control

This research proposes a generic decision making system structure for real time despatch control in small Flexible Manufacturing Systems. This is to satisfy the requirement for low cost control systems that can be flexibly adapted to a wide range of production environments. A simulation environment has been developed to emulate the detail real time despatch control of flexible manufacturing systems. This environment allows analysis of the decision making process and its effects. A model of a modular type FMS is used to study decision making in real time FMS control. Real time control is dynamic, the decision criteria change with the production states of the system. Decision making is based on both quantitative and qualitative factors. Apart from production quantity and time which are quantitative, there are installation dependent and production situations better expressed in states which are non-numeric. The knowledge based representation developed from artificial intelligence work is superior in modelling both mathematical scheduling research and discrete states information. Recognising the importance of system particular knowledge to effective control of the system, system independent functions are separated out to form elements of a generic control system architecture. This generic architecture contains elements of information handling to process information to service the scheduling decision making element. A core for regulating information flow and a data interface definintion allows this control architecture to be hardware independent. The decision making mechanism dependent on machinery hardware configuration and particular production characteristics can then be designed and interface to the architecture to form a complete control system. A decision design methodology has been designed to guide the design of the scheduling decision making system. The methodology addresses the design of work queue formation timing and the characteristics for each resource in the system. These are then integrated into a complete work flow control system by the resolution of contentions between the individual queues. The application of the design methodology and control system architecture is illustrated.