STUDIES IN THE OPTIMAL DESIGN OF FLEXIBLE CHEMICAL PLANTS.

摘要

The problem of optimal design of flexible chemical plants can be classified into two types--the deterministic multiperiod design problem and the problem of optimal design with uncertainty in the values of some of the process parameters. In both these types of problems the main objective is to design a chemical plant that has the required flexibility to meet the specifications for the various conditions that may be encountered, while being optimal with respect to a selected economic index.; An efficient decomposition scheme based on a projection-restriction strategy is developed, for solving multiperiod design problems that are formulated as nonlinear programs with block-diagonal structure in constraints, resulting in significant gains in computational effort. A rigorous mathematical formulation is presented for the problem of optimal design with uncertainty in parameter values, for which its interpretation and properties are discussed. A strategy for solving the problem of design under uncertainty is presented, which involves the solution of a series of deterministic multiperiod design problems in an efficient iterative scheme. The proposed methods result in designs that can be guaranteed to be both feasible and optimal. The application of the proposed solution algorithms and their performance are illustrated by example problems.