Cost effective and on-time delivery is required for success in today's increasingly competitive manufacturing environment. This research deals with a unified control system for distributed scheduling at the shop floor level and distributed machine capacity control at the equipment level to address these needs. Modeling, analysis and synthesis of a control-theoretic approach for scheduling using feedback control with highly distributed structure is provided, which is called distributed arrival time control (DATC). Vector space model for DATC scheduling single machine systems has been developed in order to increase the predictability of system dynamics and improve the global scheduling performance. Geometric insights gained from the vector space model have been used to synthesize control system gains to improve scheduling performance of DATC and to derive several dominance conditions for the Branch and Bound (B&B) algorithm. When production demands exceed the available resource capacity, DATC system exhibits “chattering.” Chattering is used for search mechanism to discover near optimal schedules and has been analyzed using sliding mode theory in order to increase the predictability of DATC system. A distributed machine capacity control system (DMCC) has been also developed for autonomously controlling processing parameters such as cutting speed and feed rate in real-time. DMCC has been unified with DATC in order to simultaneously adapt to changes in production scheduling demands and machine health constraints measured through real-time sensor signals. A highly distributed and time-scaled simulation has been developed for a real-time feedback of this unified control system.; Computational experiment for various problem sets has shown that DATC with the synthesized control gains can produce near optimal solutions in single machine systems, which makes it a promising alternative to conventional combinatorial optimization techniques. Numerical simulation results have shown that the unified system of DATC and DMCC can be highly adaptive in which all machines can continuously adjust part processing times without explicit notification about the health of other machines. The ability of the unified control systems to adapt to changes in production demand and changes in health of other machines for reducing total global production cost of multiple parts on multiple machines in a just-in-time manner has been demonstrated. In the long term, manufacturing control systems can improve their competitiveness by optimally utilizing capacity while reducing costs, production delays, missed deliveries, and customer dissatisfaction with the unified system of this research.
展开▼
