Modeling And Optimization Of MI Prediction For Propylene Polvmerization Process Based On Artificial Intelligent Optimization

Polypropylene (PP) is playing a more and more important role in industry, military, social life and so on, which makes the quality control of the propylene polymerization process quite crucial, especially the prediction of the melt index of polypropylene product. In this paper, the soft sensor modeling of MI by Least Squares Support Vector Machines (LSSVM) and Wavelet Neural Network (WNN) is discussed, model, and the artificial intelligent algorithms are used to optimize the structure of the model; Several improved algorithms are used to improve the model’s performance. These models established work successfully on the actual data from the practical industrial plant, thus offering more options when dealing with the melt index prediction problems.The main contributions of the present work are as follows:1、The propylene polymerization process is discussed and several input variables and several output variables are chosen to develop the model for melt index prediction, at the same time principal component analysis(PCA) is implied to simplify the input variables of model and the several measures are employed for model evaluation.2、Discuss the disadvantages of traditional FOA, and use the theory of the mutation and the change for the searching method to study the AM-FOA and IFOA. The two algorithms are employed to optimize the parameters of LSSVM and WNN. AM-FOA-LSSVM and IFOA-WNN melt index prediction models are built. Research based on the practical data shows that the AM-FOA-LSSVM and IFOA-WNN models own a very good prediction performance.3、Discuss the disadvantages of traditional FS, and use the theory of The strategy of the catastrophe to study the IFS. The improved algorithm is employed to optimize the parameters of T-S.IFS-T-S melt index prediction model is built. Research based on the practical data shows that the IFS-T-S ownss a very good prediction performance.