OPTIMAL-DESIGN OF MULTIPURPOSE BATCH PLANTS .2. A DECOMPOSITION SOLUTION STRATEGY

A mixed integer nonlinear programming (MINLP) formulation for the optimal design of a multipurpose plant has been developed in part 1 of this series. The complexity of the model makes the problem computationally intractable for direct solution using existing MINLP solution techniques. Consequently, a decomposition strategy is presented that alternately solves a MILP master problem, which determines the values of the binary assignment variables for fixed compaign lengths, and a NLP subproblem, which performs equipment sizing and determines the values of the campaign lengths. The effectiveness of the decomposition procedure is demonstrated with a number of test problems that were solved in reasonable computation times. © 1990, American Chemical Society. All rights reserved.