Research On Modeling And Quality Control Method For Vinyl Chloride Polymerization Process

Due to the strong non-linearity and complexity of mechanisms, the online monitoring, optimization and quality control of the polymerization process are challenging. This paper focuses on the online soft measurement of process parameters and quality index, the optimization of composite initiators and the optimization control of quality index for the vinyl chloride suspension polymerization. The main reseaches are summarized as following:(1) The mechanism of vinyl chloride(VC) suspension polymerization is comprehensively studied. Three basic models(the polymerization reaction kinetic model, the reaction kettle thermodynamic model and the particle size population balance model) of VC polymerization are provided, giving necessary conditions for the soft measurement, optimization and qualiy control of polymerization process.(2) A hybrid modeling method based on model output correction method is proposed. By analyzing the mechanism of vinyl chloride suspension polymerization process, a generalized population balance equation(PBE) model is developed. Then a BP neural network is constructed for estimating the error of the PBE model and applied to correct the output of PBE model. Thus, the hybrid modeling is constructed, which not only satisfies the mechanism of the evolution of the particle size distribution of polyvinyl chloride, but also take full advantage of the field data. The application of the hybrid model in predicting the particle size distribution of a practical industrial vinyl chloride polymerization process verifies that the hybrid method is effective in comparison with the pure dynamic mechanism model.(3) A multi-model fusion modeling method based on improved Kalman filter algorithm is studied. The K2PCA-ANN model and thermodynamic mechanism model are applied to estimate the vinyl chloride polymerization rate. The outputs of the two models are then fused through the recursive framework of Kalman filtering algorithm which based on the optimal estimation theoretics. The parameters of the Kalman filtering algorithm are updated through calculating the model variance to optimize the estimation performance of fusion models. The application of this multi-model fusion modeling method verifies that it is able to take full advantages of the K2PCA-ANN model and the thermodynamic model to extraction information from the field data, enhances the anti-interference ability of each model, and improves the precision of fusion models.(4) The online soft measurement technology of polymerization conversion based on system integration is studied. The OPC communication technology is applied in the online application of soft measurement technology. It contains a server station and an operator station which are working independently. The server station connects to the control system, reads the online data, computes the target variable value through the soft measurement model and writes the values into the control system. Then the results will be submitted by the OPC server immediately. The operator station connects to server station through the industrial ethernet, and the OPC client makes communications with the OPC server, thus the operator station is able to monitor the target variable. The application of monitoring VC polymerization conversion verifies that it solves the communication problem between the softwares and hardwares of different manufacturers, the computing software and configuration software have complementary advantages.(5) For optimizing the concentration of composite initiators in VC polymerization process, a self-adjustment optimal strategy combined with the thermodynamic model,kinetic model and composite initiators optimization model is proposed. The optimization target is minimizing polymerization time, constraint conditions are the the initiator concentration limitation and the heat removal capacity of the reactor, the nonlinear programming method is used to solve the optimization problem of composite initiator system. Furthermore, the influence of reactor heat transfer coefficient migration on the optimization results is fully considered, an optimization strategy is presented. Results show that this strategy is able to optimize the concentration of composite initiators, and adapt to the changes in the cooling water temperature, polymerization temperature and heat transfer coefficient to ensure the optimization effect.(6) For the quality index control of VC polymerization process, a method based on the soft measurement technology is presented. A data-driven model based on K2PLS-ANN is proposed to predict the quantitative information between the process variables and the mean particle diameter. The optimal operational variable with quality index constraint is obtained by soft measurement model prediction and nonlinear programming. According to the similarity of the batch polymerization process, a deviation elimination strategy is presented to adjust the operational variable by the offline analysis value. The application in the VC mean particle diameter control verifies that the proposed method can not only reduce the industrial polymer production cost, but also improve the stability of quality control. It is able to guide the PVC production process.On the basis of comprehensive understanding of the VC polymerization process mechanism, this paper focus on the measurement technology, optimization technology and quality control technology of polymerization process in order to enhance the ability of modeling and polymer quality control, which can be further used to guide the advanced control, system optimization of polymerization process.