Neural network based soft sensor for prediction of biopolycaprolactone molecular weight using bootstrap neural network technique

This work attempted on developing soft sensor for prediction of biopolymer molecular weight using neural network as the tool. Molecular weight is a parameter that cannot be measured online whereas it is difficult for most of us to develop and control this particular parameter. Alternatively, the molecular weight is predicted by utilizing inferential estimation method based on neural network model. In this work, temperature of biopolymerization process is used to bring a mutual relation to biopolymer molecular weight. The process involved the development of neural network model for estimation of molecular weight based on various reaction temperatures. In this study, the results are convincing and the soft sensor developed from neural network is really reliable in forecasting the biopolymer molecular weight.