A New Pressure Swing Distillation Process And Research On Operation Optimization Based On PS-PSO-GA Algorithm

A large number of azeotropic or near boiling mixtures will be produced in actual industrial production.In view of the high energy consumption of the current pressure swing distillation process of azeotrope systems,this paper integrates the characteristics of heat pump distillation,pressure swing thermally coupled distillation,and side-stream distillation,and proposes a new pressure swing distillation process.For minimum-boiling azeotrope,the characteristics are as follows: 1)Part of the overhead vapor from high pressure column is compressed as its reboiler heat source and part of it is connected in parallel as the reboiler heat source of the low pressure column;2)Cancel the condensing system of the low pressure column,and send overhead vapor to the middle of the high pressure column to upgrade the use of the waste heat;3)The liquid side-stream is extracted from the middle of the high pressure column and sent to the top of the low pressure column as reflux,forming a side-stream distillation process to reduce backmixing.For maximum-boiling azeotrope,the characteristics are as follows: 1)the two reboilers at the bottom of the low pressure column respectively use the overhead vapor from high-pressure column and the compressed overhead vapor from the low pressure column as heat sources;2)the gas side-stream is extracted from the middle of the low pressure column,and the heat is supplied to the bottom of the high pressure column after compression and correspondingly its reboiler is cancelled;3)The bottom stream of the high pressure column is sent to the middle of the low pressure column,forming a side-stream distillation process to reduce backmixing.The Pro/II calculation results of the four minimum-boiling azeotrope systems of phenyl-ethanol,methanol-acetone,ethyl acetate-methanol,and ethyl acetate-ethanol and a maximum-boiling azeotrope system of ethylenediamine-water indicate that comparing the complete thermal integrated pressure swing distillation process,the annual total cost of the new process is reduced by 31.26% on average and the thermodynamic efficiency is increased by 17.50% on average.Then,the robustness of process unit inner mathematical model in simulated-data-driven optimization(SDDO)is studied,and the method for robust optimization of SDDO model is proposed.the application of an atmospheric crude oil distillation column shows that the convergence scenario increases by 15.03%,which greatly reduces the influence of external perturbation on the process simulation,such as the selection and value of adjustable operation variables,system product quality constraints,etc.It lays a foundation for the subsequent operation optimization research.Furthermore,a new constraint processing method is proposed,which is a hybrid method combining the death penalty function method and the feasibility rules;A new PS-PSO-GA optimization algorithm is proposed by integrating the characteristics of strong global search ability of genetic algorithm(GA),high local optimization accuracy of pattern search(PS),and the fast convergence of particle swarm Optimization(PSO).On the basis of the above,a new SDDO problem system method based on the integration of PRO/Ⅱ and Excel software has been developed.The new system method is applied to the new pressure swing distillation process developed in this paper.Taking maximum-boiling azeotrope ethylenediamine-water system as an example,the total operating cost is reduced by 30.41%compared to the design conditions,and it has been reduced to 28.69% in the 27 th generation,which is much higher than the 11.79% that GA converges in 44 th generations when optimized alone.It indicates that the SDDO system method proposed in this paper is suitable for the multi-variable and strongly constrained operation optimization research of all chemical units that can be simulated.