A NEW OPTIMIZATION STRATEGY FOR CHEMICAL PROCESS DESIGN UNDER UNCERTAINTY (TWO-STAGE, QUASI-NEWTON, HAN-POWELL)

The formulation of two-stage decision problems in the application to chemical process design with uncertain parameters is investigated and a new optimization strategy is proposed for solving these problems.; A systematical procedure is found for obtaining all the two-stage problems for a given design situation; examples are given to illustrate the application of this procedure.; The proposed optimization scheme combines gradient-based nonlinear programming routines with experimental design and statistical testing techniques. The Han-Powell algorithm is particularly adapted for this purpose, and its efficiency is illustrated through solving algebraic problems and process problems modeled by a process simulation package.