A control-theoretic approach to production planning and control of a multi-workstation production system

Production Planning and Control (PPC) has become more challenging as manufacturing companies adapt to a fast-changing market. This turbulent environment requires the companies to plan and control their production in such a way that the disruptions in performance of their production systems are minimized in order to remain competitive. Current PPC methods, which are based on a static view of the production systems such as infinite capacity and fixed lead time, are growing increasingly inadequate in these dynamic situations. New PPC methods need to be developed that facilitate rapid response to turbulent conditions with dynamic properties that are well-behaved and robust.; An approach for closed-loop Production Planning and Control (PPC) for capacity control of a multi-workstation production system is developed in the dissertation, that adjusts capacities of individual workstations to regulate WIP levels at the workstations. The dynamic of the multi-workstation PPC system are represented by transfer equations that relate system inputs such as work and capacity disturbances to performance-related outputs such as output rates and WIP deviations with respect to planned WIP levels. An order flow matrix represents order flow relationships between workstations. The effects of delays in capacity adjustment and adding order flow loops between workstations are investigated along with the limitations they place on controller gain.; Using a six-workstation production system, it is shown that locally applied capacity control alone is ineffective to eliminate upstream disturbance effects, but adding capacity adjustment coupling, reflecting communications between workstations, enables individual workstations to predict and eliminate the propagating effects. The potential behavior of an industrial system incorporating the closed-loop PPC for capacity is investigated using a set of production data from a manufacturer of automotive components.