Principle Of Full Cycle Dynamic Optimization For Slowly-time-varying Chemical Processes

In a long-operating-cycle chemical process system,there are parameters that change slowly as time goes on,such as fouling thermal resistance of the heat exchanger network and catalyst activity of the acetylene hydrogenation reactor.The slowly-time-varying parameters can be treated as parameters in the short-term operation of the system,and their variation can be ignored.However,in the long period of operation,its influence on the whole system cannot be ignored,so it should be treated as a variable.With acetylene hydrogenation reactor for a case,this paper discuss three aspects for the full cycle slowly-time-varying systems optimization problem,i.e.full cycle steady-state optimization,dynamic modeling,and full cycle dynamic optimization.On this basis,in this work,a margin of slow-release full cycle operation optimization method is proposed.The sufficient theoretical proofs are given for its slow-release characteristics.First of all,a novel two-dimensional heterogeneous modeling approach of acetylene hydrogenation reactor with catalyst deactivation model is proposed,and the model is employed for the discussion of a commercial acetylene hydrogenation reactor unit.Considering the cumulative effects of secondary reaction on catalyst activity,the more strict catalyst deactivation model that is different from the previous research on modeling approaches is proposed for accurate estimation of catalyst activity,and the complete acetylene hydrogenation reactor dynamic mathematical model is established.Next,for the full cycle optimization of the reactor,the time scale difference between the control system and the optimizer is a notable problem.If the optimization of the two systems is carried out together,it will cause a large computational redundancy.Hence,a full cycle optimization framework combining fast-time-varying system and slowly-time-varying system is proposed.According to the framework,the steady-state optimization model whose objective is diurnal economic benefit is constructed,and the solution is compared with the manual temperature compensation effect in actual production according to the worker experience.Furthermore,based on the framework,the full cycle dynamic optimization of slowly-time-varying systems is discussed.According to the characteristics of the acetylene hydrogenation reactor model,an optimal model of integration economic benefits within full cycle is established within a certain regeneration cycle.Considering the control effect in the fast-time-varying system,the controller parameters should be solved together with the optimization variables.In addition,sometimes,it is necessary to extend the regeneration cycle of the unit as far as possible for the factory scheduling.Thus,an optimization model whose objective function is the maximum regeneration cycle is proposed in this work,which is solved and compared under the same optimization framework.Moreover,an on-line catalyst activity estimator is designed,which can effectively estimate the activity of catalyst in the reactor.In the actual industry,especially the system with a long operating cycle,considering the influence of many interfering factors,design margin should be reserved in the design stage to ensure the long-term stable operation of the system.However,as part of the design margin,process margin is related to the variation of slowly-time-varying parameters of the system,it increases slowly in a regeneration cycle.Considering that surpuls margin can be effectively utilized before the required process margin reaches its maximum value,a full cycle dynamic optimization method maintaining the operation margin is proposed.The method is implemented in the optimization framework of slowly-time-varying system.Furthermore,based on the minimum principle,the characteristic of the margin slow-release in slowly-time-varying systems is proved,and the general applicability of the full cycle dynamic optimization method maintaining the operation margin in the slowiy-time-varying systems is illustrated.In the last part of this paper,the whole content of this paper is summarized,and provides the direction of future follow-up work.