Research On Control Of Molecular Weight Distribution In Styrene Polymerization Via Event Triggering

With the rapid development of the new energy industry,higher requirements are placed on the production process of styrene and its chemical derivatives,mainly including the physical and chemical properties of the products as well as reasonable cost control.Among them,the molecular weight distribution(MWD)of styrene is an important indicator of its physical and chemical properties.Therefore,the study of modeling,control and optimization of styrene polymerization process has received a lot of attention in the field of process control and chemical industry,and has important practical applications.On the other hand,due to the complexity of the chemical process,especially multiple information sources generate a large amount of transmission data,which excessively occupy channel resources,causing problems such as data redundancy and system performance degradation,and even loss of stability in severe cases.Compared with the traditional time triggering,the basic idea of event triggering mechanism is to execute data transmission on demand,and the trigger is activated only when the preset triggering conditions are met,which is one of the reliable solutions to reduce data redundancy at present and has become a hot spot for research in the field of automatic control.Based on this,the theoretical framework of B-sample approximation and statistical information control,combined with the interference observer design method,the paper investigates the dynamic modeling,anti-interference control and performance analysis of styrene molecular weight distribution in the event-triggered framework to ensure the stability and robustness of styrene polymerization process,reduce channel resource utilization and thus achieve cost control.It is of positive theoretical significance and application value to achieve cost control.The main contents of the paper include the following:(1)The research background and significance of molecular weight distribution control under event-triggered conditions are described for a typical styrene polymerization process.Then,the current research status of random distribution control(SDC),statistical information control,anti-disturbance control and event-triggered control are discussed respectively.Further,the structure and content arrangement of the paper are described.(2)The reaction mechanism of the styrene polymerization process and the distribution function modeling problem are studied.Based on the parametric mechanistic equations of the polymerization process.a previous distribution function,i.e.,a three-parameter Gamma distribution.is introduced.On the one hand.driven by the B-sample approximation model,the data are transformed into corresponding data pairs of weight vectors and control inputs,and the system parameter matrix is obtained by recursive least squares identification,and then the dynamic model between the control inputs and the weight vectors is established;on the other hand,in the framework of statistical information study,the three central moments of the threeparameter Gamma distribution constitute the statistical information set(SIS),and then the statistical information set is obtained with the data pairs of control inputs for identifying the parameter matrices.which can likewise build a dynamic model between the control inputs and the statistical information set.(3)To study the control problem of interference resistance of molecular weight distribution of styrene polymerization process under event-triggered conditions.Based on the dynamic model between the control input and the power vector,an interference observer is constructed to estimate the unknown interference dynamically by combining the episodic interference model.Based on the predefined event trigger conditions as well as the disturbance estimation values,the composite anti-disturbance control input under the trigger conditions is designed to ensure the stability and disturbance estimation performance of the closed-loop control system on the basis of reducing the data transmission.Using convex optimization theory,it is demonstrated that the weight vector has good dynamic tracking performance,and then the desired molecular weight distribution is obtained.Finally,it is verified that the designed algorithm avoids the Zeno phenomenon caused by triggering.(4)To study the problem of statistical information immunity control of styrene aggregation process under dual-event triggering conditions.Based on a data-driven statistical information control framework,the statistical information set is used as the specific control object.In contrast to the single-event trigger problem,the trigger conditions are designed based on statistical information and disturbance estimation,and the anti-disturbance control input with PI structure is designed based on the "first-come-first-trigger" operation mechanism to realize the control and disturbance compensation of the dynamic system.Further,based on the convex optimization theory and the proof of the lower bound of the minimum trigger interval,it is verified that the closed-loop system has good stability,statistical information tracking performance and the avoidable Zeno phenomenon.(5)The simulation implementation of the algorithm is investigated.Based on two dynamic models for modeling the styrene aggregation process,single-trigger and doubletrigger conditions are considered,and the controller and its related algorithms are simulated with the Matlab/Simulink dynamic simulation platform and the LMI toolbox,and the simulation results verify the feasibility of the designed algorithms.