Nested partitioning algorithm for the optimization of control parameters in energy efficient production lines

Energy efficient control policies that switch off/on the machine using buffer occupancy information have been recently proposed in literature, considering that machines may need a transitory before resuming the service. However, how a simultaneous control of the machines may affect system performance due to the propagation of blocking and starvation effects is not trivial. The optimization of control policies parameters requires system performance evaluation through simulation, but it is highly time consuming as system complexity increases. This work aims to design an efficient Nested Partitioning algorithm for the optimization of switching control policies in production lines based on the structural properties that the optimal control might have at system level. The performance of the algorithm will be evaluated using discrete event simulation and compared with a commercial optimization tool (OptQuest). The effect of the control on the system throughput will also be reported.