摘要
一、引言一套印处理量为250万吨的炼油常减压装置由初馏塔、常压塔、减压塔和加热炉组成。它有八个侧线,产品为汽油、柴油和润滑油。优化控制目标为提高收率和保证产品质量(干点,凝固点,350℃馏出,粘度)。为此,本文提出了“卡边温度”概念:使产品质量控制在靠近工艺上限的指定区间内(叫质量卡边域)的侧线平均馏出温度叫“卡边温度”。由蒸馏机理知质量卡边有利于提高收率。于是问题转化为对各侧线按其卡边温度控制问题。本文提出一种基于模式识别原理在线计算各侧线的卡边温度的方法。由蒸馏机理知,在采样时刻τ,侧线产品质量化验值Z(τ)主要由塔压力P(τ-k)和侧线馏出温度T(τ-k)决定,k是采样滞后时间,称(Z(τ),P(τ-k),T(τ-k))为样本,称所有样本集合Q为样本空间。设在目前时刻t塔压力为P(t),则视Q中使Z(τ_i)
Control of large crude oil distillation unit according to bordering temperature of sidelines was studied. To remove strong coupling between temperature of different sidelines, multivariable decoupling fuzzy control with double-layer structure was proposed. The upper layer includes a fuzzy coordinator while the lower layer consists of 12 single-channel fuzzy controllers. Control results for a distillation unit show its superiority over manual regulation.
出处
《自动化学报》
EI
CSCD
北大核心
1993年第6期750-752,共3页
Acta Automatica Sinica
关键词
石油炼制
常减压装置
解耦
multivariable fuzzy control systems
decoupling
crude oil distillation
computer process control.
