A short-cut method for optimal selection of controlled variables

摘要

Selecting an optimal set of controlled variables requires the evaluation of all possible sets of controlled variables. The number of possible sets of controlled variables grows exponentially with the number of candidate controlled variables. The number of candidate controlled variables is generally much alrger than the number of controlled variables that can be chosen. This implies that finding an optimal set of controlled variables involves solving a large number of non-convex optimization problems, a computationally intensive task for any realistic chemical process. A short-cut method, which generates attractive sets of controlled variables and determines an optimal set, is proposed. The method is based on scaling all the candidate controlled variables so that they have similar effects on a profit function. The procedure is illustrated on a butane alkylation process.