Study On Optimization Under Uncertainty For Chemical Seperation Process

Over the second of half of the 20th century, optimization found widespread applications in the study of physical and chemical systems. From the very beginning of the application of optimization, it was recognized that uncertainty can not be avoided during optimization. Normally, uncertain parameters are treated by its overestimation values to avoid infeasibilities, which would cause much more costs than necessary. Therefore, uncertainty should be considered in the optimization.The optimization strategies are studied for the chemical separation system in this dissertation, and a hybrid process integration method is presented for the optimization under uncertainty based on stochastic strategy.The mainly content of this paper is listed as follows:(1) The source and classification of the uncertainty for the chemical separation system are studied firstly. And then, a hybrid stochastic optimization strategy is proposed for the chemical separation process under uncertainty. The approach of two-stage stochastic programming with recourse is utilized for the economic objective, and the detailed recourse expression is presented in this paper. The approach of chance constrains is suitable for the environmental objective considering that some constraints can be violated with a certain probability p∈(0, 1) while the system is thought to satisfy the corresponding constraints. The efficacy of this methodology is demonstrated by the product separation process optimization in the production of 1-hexene.(2) For the solution on the model of two-stage stochastic programming with recourse, this paper attempts to extend the prior research in this field to develop an approximated integration algorithm with Monte Carlo integration and improved Benders decomposition strategies. To overcome the great difficulty of the calculation that lies in the great deal of constraints caused by sampling in the whole region, this paper proposes to delay the constraints generated through positive constraints to reduce the calculating duty. This improved algorithm shows more realistic and effectiveness through the problems test than the former work. What is more important, it can deal with the large-scale stochastic programming problem.(3) The case study about the separation process optimization in the production of 1 -hexene illustrates the performance of the hybrid stochastic programming strategy presented in this paper. Three uncertainty parameters are considered, and they are the variation of LP/MP steam price, the application of new catalyser and the production throughput enlargement, and two objectives both of economic maximum and environmenatal impact minimum are optimized simultaneously. The study shows that the optimum alternative and decision factors are different by optimization under uncertainty compared with that of the deterministic optimization for the case study.From the work, it is obviously known that it is necessary to take the uncertainty into consideration in the chemical separation process optimization.