Ethylene Unit Cracking Furnace Predictive Control And Optimize Based On Model

The product quantity and quality of etibylene production device not only is directly related with economic benefits but also influences the normal production movement and production quality of large downstreem production device (such as polyethylene device,polypropylene device,glycol device),as a result it accuplies a outstanding position in the quality (yield)control of enthylene production process. This text estabishes the pridict model of ethylene device cracking furnace according to orthogonal design method and stepwise regression method to predict control and optimization.Firstly, this text introduces mathematics model of the craft mechanism of the estabishment cracking furnace and then estabishes ethylene device cracking furnace's simplified model according to orthogonal design method and stepwise regression method.The experiment deta is made use by the method of nonlinear regression method to get the simplified computer controlling model of ethlene and propylene.The analysis of variance and stastistics test of regression equation and regression coefifcient suggests the regression results is good ,the results of mechanism model and simplified model are very similiar and the exact rate of simplified model is very high,as a result this model can be made use to replace mechanism model which is used for computer control.Secondly,this test mainly introduces the soft measurement of the cracking furnace which has something with variables of cracking depth,the outlet temperature of cracking furnance,the production ability of cracking furnance and the raw materials ,and the steam dilution ratio control,in order to improve the yield of ethylene.In the end ,the cracking furnace gets optimal operated further on the basis of advanced control the ethylene device cracking furnace,in order to get the greatest average yield.This chaper establishes optimal proposition on the basis of constant conditions,applied network for fast training methods and algorithms ,decribes the operating characteristic idustrial pyrolysis process exactly and uses the optimal algorithms of simulated annealing for optimizing system.