Ethylene Cracking Furnace's Soft Measurement And Operation Optimization Research

Natrual gas, refinery gas, crude oil, naphtha and other kinds of raw materials are processed into cracking gas by cracking furnace unit, which is the core equipment of ethylene production. After cracking gas transferred to the remaining devices to get final products, such as, ethylene, propylene and various kinds of by-products. At present, the product quality needs artificial laboratory analysis. There are corresponding analysis for a long time and measurement lag problem. The online chromatograph is too expensive to maintain, so it's hard to popularize. So we encounter the problems of inconvenience and hysteresis. To aim at this problem, this paper put forward the development of ethylene cracking furnace of ethylene, propylene yield of soft measurement technology is used to guide the production operation.At the same time, the technology and theory on soft sensor are studied, including the distributed control system (DCS), data communication technology, etc. This paper also compares the Back Propagation (BP) neural network algorithm with the Radial Basis Function (RBF) neural network algorithm by network training and testing, and realizes the soft sensor for ethylene cracking furnace of ethylene and propylene yield on the basis of RBF neural network algorithm is better. In this paper, a new type of swarm intelligence optimization is studied. Its optimization process needs to obtain the new luciferin number, to find out the best ones, to calculate the new location and to divide the perception range. The GSO method is used in ethylene cracking furnace, to get the device operating parameters of the steady-state optimization problem for the optimization, improving double-ene yield, to make economic benefits for enterprise security, stability, sustainable, efficient.