Font Size: a A A

Intelligent Modeling And Optimization For Polymerization System Of High Density Polyethylene Unit

Posted on:2011-10-11Degree:MasterType:Thesis
Country:ChinaCandidate:N LangFull Text:PDF
GTID:2121360305985100Subject:Control Science and Engineering
Abstract/Summary:PDF Full Text Request
High density polyethylene is a high crystallinity and non-polar thermoplastic resin produced by copolymerization of Ethylene. It has excellent chemical properties, and has been widely used as the main materials of film, plastic and pipe in industry and daily life. Hostalen low pressure slurry technology is the world's leading technology of the production of high density polyethylene, using two parallel or series reactors for slurry polymerization. Polymerization process is the core of the polyethylene production process. It has great significance for improving product quality, reducing the unit consumption and getting the greatest economic benefit through holding polymerization parameters and determining the optimal operation of the program. However, the mechanism of high density polyethylene industrial process is very complicated and not perfect, and some parameters must be obtained through experimental methods, which are difficult to achieve in the absence of experimental conditions. In addition, a number of online analyzers could not be used, and brings inconvenience for monitoring production status timely, it also brought difficulties to the online optimization, besides, the technology don't have referenced operation experience.Based on the above problems existing, this paper studies the method of modeling by using neural network on the basis of the analysis of internal polymerization mechanism of high density polyethylene, and have modeling to the tandem operation of polymerization process combined with the practical problems in industry, including product quality-melt index and density soft-sensing model. This paper provides a method based on sensitivity analysis, aiming to confirm the input variable set and the number of hidden layer nodes more accurately. Industrial application result has fully proved the effectiveness of the method.The following study concludes the establishment of optimization model for the target of quality controlled and energy saved, and the optimizing operation using PSO optimizing technology, seeking the best operating plan to reduce the unit consumption. It has great significance for improving product quality, saving production cost and increasing economic benefit.
Keywords/Search Tags:high density polyethylene, neural network, soft sensing, sensitivity analysis, input variable selection, modeling, optimization
PDF Full Text Request
Related items