Optimization Of Distillation Process For Close-Boiling Methylchlorosilane Monomers With Intelligent Optimization Algorithms

Methylchlorosilanes is an important synthetic monomer for a variety of silicone materials.In its production process,the energy consumption of the distillation unit occupies about 70% of the total energy consumption.Consequently,it is of significant economic benefits to improving and optimizing this distillation process.Conventional optimization methods for the simulative distillation model frequently relies too much on human experiences to reach the global optimal solution.In this scenario,intelligent optimization algorithm based on probabilistic search can be taken into account as a solution for its low mathematical information requirements and strong global searchability.The global optimal solution becomes more approachable with this algorithm in application to optimize such distillation processes that suffer from uncertainties from empirical optimization methodologies.With these the targets for the thesis,in the first place,the Aspen/MATLAB integrated optimization procedures are programed to import Aspen distillation model calculation data into MATLAB where to be optimized by the coded intelligent algorithm.Afterwards,the differential evolution algorithm with improved search strategy is used to optimize the most energy intensive part of the popular methylchlorosilane distillation process(MDP).Under the premise of the given basic process,this algorithm optimizes its operation parameters and obtain corresponding operative temperature by all the column pressures,then inplement heat integration structures into the process.Finally,3 improved basic processes are proposed and optimized by the same intelligent algorithm,with their annualized total costs(TACd)in comparison for the best of the largest TAC reduction.The specific research results are as follows:(1)an adaptive dynamic differential evolution(SADDE)algorithm is used to optimize the operation parameters of MDP.The TAC is reduced to 15.77 × 10 $/a,16.25% lower than the recent published research.(2)Referring to stream matching strategy,a method is proposed to automatically generate the corresponding heat integration structure according to the pressure of each column.Then the proposal method is embeded into SADDE algorithm for the heat integrated SADDE(HISADDE)algorithm.By simultaneous optimization/screening of process para-meters and heat integration layout through HISADDE,The MDP is further optimized to the heat integrated MDP,with an optimized TAC of13.78 × 10 $/a,lowered by 12.59%.(3)Propose 3 new MDP schemes by splitting the distillation unit of monomethyltrichlorosilane and dimethyldichlorosilane,namely parallel-flow,counter-flow and forward-flow processes(PFP,CFP and FFP).Then use HISADDE algorithm to optimize the 3 processes.FFP with cascaded heat integrated structure performs best,optimizing TAC to 10.76 × 10 $/a,31.77% lower than the MDP,21.97% lower than the heat integrated MDP.In addition,FFP is further optimized by thermal coupled sequence,reducing TAC by another 2.3% compared with FFP.In summary,by the Aspen /MATLAB integrated optimization program upon HISADDE algorithm,the intelligent optimization is readlized of not only the operation parameters,but also corresponding heat integration structure for a given basic process at the same time,This approach enlarges the calculation content the optimization program from proassigned scheme to autonomous choice of heat integration for it.Its applicability and globality is proved by optimizations of MDP processes of various complexity,which results in,after analyses,key technical points for such heat integrated MDP.The way of thinking and the methodology in this thesis can be popularized as reference for similar complicated distillation processes.