Integration of simulated annealing to a simulation tool for dynamic optimization of chemical processes

Abstract

Dynamic optimization of processes in chemical industry became important in the last decade. Simulated annealing belongs to these optimization procedures, which allow to reach the global optimum, but are very time-consuming. In this paper, the simulated annealing algorithm has been formulated in a way to enable direct call of process simulators coded in the commercial simulation tool SPEEDUP. We apply simulated annealing for the determination of operation policies in a binary distillation. The developed approach allows the prediction of reflux flow rate and reboiler duty in order to minimize the operation time for product switchover of the column.