Product Design And Transfer Using Data And Lantent Variable Modeling Methods

In the industrial production, process can often be operated over a range of conditions to produce a variety of different grades of a product to meet the requirements of the market and customers. In order to achieve certain economic specification and product quality requirement, operating variables often need to be adjusted correspondingly. Consequently, how to determine a window of operating conditions within which one should be able to produce a product having a specified set of quality characteristics,especially if the cost of raw materials is high or the product manufacturing is subject to a rigid regulatory environment. If a method is developed to define a set of operating conditions to produce a desired product, it would effectively shorten production cycle,reduce production cost, thereby enhancing production efficiency. Obviously, the study of product design and its application in the industrial production has great theoretical and practical significance.At the beginning, this thesis makes an analysis of product design problem theoretically. Then, the importance of estimating the operating conditions in a target plant through using the latent variable modeling approach is analysed. Based on intensive study, methodologies using historical data and latent variable models to solve product design problem are summed up and summarized, developing a common framework of product design theoretically in this paper.The application of designing process conditions through latent variable modeling to a simulated LDPE polymerization process verifies the effectiveness of the method. As for product design for nonlinear case, the main features of previously introduced linear model inversion are therefore extended to nonlinear inversion, several issues and cautions involved in the use of nonlinear models and techniques versus linear ones are raised at last.As for product design of a process with insufficient data information,this thesis aims at study the product migration strategy for processes with some type of similarity, developing two new methods:Joint-Y PLS(JYPLS) and Extended PCR(EPCR), which take advantage of historical process data on both plants to solve the product design problem of the new process.Simulation results verify the effectiveness of these two methods, and investigate their application categories in solving product design problem.