On-line optimization of batch polymerization processes

This paper deals with the issue of on-line optimization of batch processes. Batch processes are transient in nature; the process variables undergo significant change during the batch cycle and there is no process steady state. The objective is not to keep the system at a set point, but to optimize an objective function (such as yield) at the end of the batch cycle. The optimal operating strategy for batch processes involves the calculation of an optimal state trajectory (as opposed to an optimal steady-state in continuous process). In commercial batch operations the initial conditions of the batch as well as the process parameters can vary from batch to batch. Due to this change, the optimal operating trajectory also changes from batch to batch and so it is necessary to compute the optimal operating trajectory on-line. This is the main focus of this paper. In particular, we will consider a batch polymerization of methyl methacrylate and demonstrate via simulations, an on-line optimization strategy for end-point optimization.