Real-time hybrid simulation of manufacturing systems for performance analysis and control

Abstract

Simulation is a common way to analyze the performance of a manufacturing system. On a system level, entities are modeled based on production and failure rates using Discrete Event Simulation (DES). On a machine level, the Continuous Dynamics (CD) of machines are studied based on position, velocity, and acceleration. However, these models are usually run separately, losing key information that could be used for more accurate control of the system performance. In this paper, we merge DES and CD in a single hybrid simulation environment. By synchronizing the simulations to run in realtime, the results can be compared with information from the plant floor. Using the simulation outcome as a reference, any significant deviation of the plant floor performance would represent an error and would trigger an event that can be automated or inform an operator of a required action.